---
base_model: coder3101/Qwen3.5-4B-heretic
base_model_relation: quantized
quantized_by: DBMe
library_name: exllamav3
pipeline_tag: text-generation
license: apache-2.0
tags:
- exl3
---

# DBMe/Qwen3.5-4B-heretic-exl3

EXL3 (ExLlamaV3) quantizations of [coder3101/Qwen3.5-4B-heretic](https://huggingface.co/coder3101/Qwen3.5-4B-heretic). All credit for the original model goes to the original authors.

## 📊 Available Quantizations & VRAM
The model weights are stored in separate branches. **Please switch to a branch to download.**
*Note: VRAM estimates include PyTorch context overhead (~0.8GB) and assume an unquantized FP16 KV cache.*

| Target BPW | Head BPW | Branch (Download Link) | WikiText-2 PPL (512 ctx)¹ | 2K ctx | 4K ctx | 8K ctx | 16K ctx | 32K ctx |
|---|---|---|---|---|---|---|---|---|
| 4.0 | h6 | [4.0bpw_h6](https://huggingface.co/DBMe/Qwen3.5-4B-heretic-exl3/tree/4.0bpw_h6) | 10.2665 | ~4.83 GB | ~4.9 GB | ~5.02 GB | ~5.27 GB | ~5.77 GB |
| 5.0 | h6 | [5.0bpw_h6](https://huggingface.co/DBMe/Qwen3.5-4B-heretic-exl3/tree/5.0bpw_h6) | 10.1381 | ~5.25 GB | ~5.31 GB | ~5.44 GB | ~5.69 GB | ~6.19 GB |
| 6.0 | h6 | [6.0bpw_h6](https://huggingface.co/DBMe/Qwen3.5-4B-heretic-exl3/tree/6.0bpw_h6) | 10.1020 | ~5.66 GB | ~5.73 GB | ~5.85 GB | ~6.1 GB | ~6.6 GB |
| 8.0 | h8 | [8.0bpw_h8](https://huggingface.co/DBMe/Qwen3.5-4B-heretic-exl3/tree/8.0bpw_h8) | 10.1099 | ~6.64 GB | ~6.7 GB | ~6.83 GB | ~7.08 GB | ~7.58 GB |

¹ *Evaluated against WikiText-2 with ExLlamaV3 using a strided 512-token context window (-c 512) in llama.cpp parity mode (-g). Lower is better.*
*(Higher BPW = higher quality, lower BPW = fits in less VRAM).*

## 📥 How to Download
It's recommended to use the `huggingface-cli` to download specific branches. *(Do not use `git clone` as it will download all branches!)*

Ensure you have the CLI installed:
```bash
pip install -U "huggingface_hub[cli]"
```

Download a specific branch (e.g., `4.0bpw_h6`):
```bash
# Example: Downloading the 4.0bpw_h6 branch
huggingface-cli download DBMe/Qwen3.5-4B-heretic-exl3 --revision 4.0bpw_h6 --local-dir Qwen3.5-4B-heretic-exl3-4.0bpw_h6
```

## 💻 Supported Engines
These models are highly optimized for modern GPUs and can be run using:
* **[TabbyAPI](https://github.com/theroyallab/tabbyAPI):** A fast, OpenAI-compatible API server. *(Set `model_name: "Qwen3.5-4B-heretic-exl3-<BranchName>"` in your config)*
* **[Text-Generation-WebUI](https://github.com/oobabooga/text-generation-webui):** A local web interface. *(Select the `exllamav3` loader)*
* **[ExLlamaV3 (Native)](https://github.com/turboderp-org/exllamav3):** Python library for custom integration.

### 📈 Perplexity Degradation Curve
*(Lower is better)*
![Perplexity Graph](https://huggingface.co/DBMe/Qwen3.5-4B-heretic-exl3/resolve/main/metrics_graph.png?v=1776316917)

<details>
<summary><b>⚙️ Advanced: Quantization Environment & Settings</b></summary>

### 🔬 Quantization Settings
- **Codebook:** mcg
- **Output Scales:** always
- **Calibration Rows:** 250
- **Calibration Cols:** 2048
- **Calibration Dataset:** ExLlamaV3 Default (Wiki/C4/Code)
- **High Quality (HQ) Mode:** False

- **ExLlamaV3:** `0.0.29` (Commit: `cb1a436`)
- **Hardware:** `NVIDIA RTX PRO 6000 Blackwell Server Edition`
</details>