---
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen2.5-Coder-0.5B-Instruct/blob/main/LICENSE
language:
- en
pipeline_tag: text-generation
base_model: Qwen/Qwen2.5-Coder-0.5B
tags:
- code
- codeqwen
- chat
- qwen
- qwen-coder
- heretic
- uncensored
- decensored
- abliterated
- reproducible
- gguf
library_name: transformers
---

# Qwen2.5-Coder-0.5B-Instruct-heretic

A decensored variant of [Qwen/Qwen2.5-Coder-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-0.5B-Instruct), produced with [Heretic](https://github.com/p-e-w/heretic) v1.4.0 (directional ablation / "abliteration"). Refusal behavior is suppressed via targeted weight edits to the attention output and MLP down-projections rather than fine-tuning, so the base model's knowledge and instruction-following are left largely intact.

**Who this is for:** developers who want Qwen's code-focused small model without the refusal guardrails — for local code agents, local copilot-style use, roleplay, research on alignment/refusal mechanics, or any use case blocked by RLHF-era over-refusal. At 0.5B parameters it runs comfortably on CPU and is ideal for on-device deployment.

## Abliteration parameters

| Parameter | Value |
| :-------- | :---: |
| **direction_index** | 19.54 |
| **attn.o_proj.max_weight** | 1.26 |
| **attn.o_proj.max_weight_position** | 14.42 |
| **attn.o_proj.min_weight** | 0.79 |
| **attn.o_proj.min_weight_distance** | 12.89 |
| **mlp.down_proj.max_weight** | 1.42 |
| **mlp.down_proj.max_weight_position** | 14.14 |
| **mlp.down_proj.min_weight** | 0.52 |
| **mlp.down_proj.min_weight_distance** | 13.14 |

## Performance

| Metric | This model | Original model ([Qwen/Qwen2.5-Coder-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-0.5B-Instruct)) |
| :----- | :--------: | :---------------------------: |
| **KL divergence** | 0.1249 | 0 *(by definition)* |
| **Refusals** | 8/100 | 52/100 |

KL divergence of 0.12 on the output distribution is low — the edit is narrow and targeted rather than a broad perturbation. Refusals dropped from 52 to 8 out of 100 adversarial prompts, meaning the model complies while retaining nearly all of its original capabilities.

> Made with ❤️ by **RACER IS OP** — follow for more uncensored models

## Files

| File | Format | Size |
|---|---|---|
| `model.safetensors` | BF16 | 988 MB |
| `Qwen2.5-Coder-0.5B-Instruct-heretic.gguf` | GGUF, F16 (unquantized) | 948 MB |
| `Qwen2.5-Coder-0.5B-Instruct-heretic-Q8_0.gguf` | GGUF, Q8_0 | 506 MB |
| `Qwen2.5-Coder-0.5B-Instruct-heretic-Q5_K_M.gguf` | GGUF, Q5_K_M | 401 MB |
| `Qwen2.5-Coder-0.5B-Instruct-heretic-Q4_K_M.gguf` | GGUF, Q4_K_M | 379 MB |

## Reproducibility

Unlike most abliteration repos, the full run is reproducible from the [`reproduce/`](https://huggingface.co/saidutta69/Qwen2.5-Coder-0.5B-Instruct-heretic/tree/main/reproduce) folder in this repo:

- `config.toml` — exact Heretic configuration used for this run
- `reproduce.json` — full parameter and metric dump
- `Qwen--Qwen2--5-Coder-0--5B-Instruct.jsonl` — evaluation transcripts against the base model
- `SHA256SUMS` — checksums for integrity verification
- `requirements.txt` — pinned environment for re-running the ablation

## Quickstart

```bash
# llama.cpp
llama serve -hf saidutta69/Qwen2.5-Coder-0.5B-Instruct-heretic
```

```python
# transformers
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "saidutta69/Qwen2.5-Coder-0.5B-Instruct-heretic"
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)

messages = [{"role": "user", "content": "Write a quick sort algorithm in Python."}]
inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=True,
                                        return_dict=True, return_tensors="pt").to(model.device)
out = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(out[0][inputs["input_ids"].shape[-1]:], skip_special_tokens=True))
```

Also runnable via Ollama, LM Studio, Jan, vLLM, SGLang.

## Responsible use

Refusal suppression is deliberate and works as intended: this model will comply with requests the base model would refuse, including some it shouldn't. There is no safety filtering layered on top. You are responsible for how you deploy it — don't put this behind an unmoderated public-facing endpoint serving third parties. It inherits Qwen2.5-Coder-0.5B-Instruct's factual limitations and biases; abliteration removes refusal directions, it doesn't add capability or judgment.

## License

Inherits the Apache 2.0 license from the base model.

## Related

- [Qwen2.5-0.5B-Instruct-heretic](https://huggingface.co/saidutta69/Qwen2.5-0.5B-Instruct-heretic)
- [Qwen2.5-1.5B-Instruct-heretic](https://huggingface.co/saidutta69/Qwen2.5-1.5B-Instruct-heretic)
- [Qwen2.5-Coder-3B-Instruct-heretic](https://huggingface.co/saidutta69/Qwen2.5-Coder-3B-Instruct-heretic)