---
license: apache-2.0
base_model:
  - orcarouter/Qwen3.8-27B-Uncensored
base_model_relation: quantized
pipeline_tag: image-text-to-text
library_name: transformers
quantized_by: writetoasik
language:
  - en
  - zh
tags:
  - exl3
  - exllamav3
  - quantized
  - qwen
  - qwen3
  - qwen3.8
  - uncensored
  - abliterated
  - tabbyapi
  - vision-language
  - function-calling
  - reasoning
  - mtp
---

# Qwen3.8-27B-Uncensored — EXL3 4.0 bpw

ExLlamaV3 (EXL3) quantization of [`orcarouter/Qwen3.8-27B-Uncensored`](https://huggingface.co/orcarouter/Qwen3.8-27B-Uncensored), which is an abliterated (refusal-removed) build of [`Qwen/Qwen3.8-27B`](https://huggingface.co/Qwen/Qwen3.8-27B).

This is a **weight-only re-quant**, not a new train. The vision tower and MTP speculative-decoding head are kept.

| | |
|---|---|
| **Format** | EXL3 safetensors (2 shards, **15.73 GB**) |
| **Bits** | 4.0 bpw body, **6-bit** output head, **4-bit** MTP |
| **Measured bitrate** | 4.02 bpw / 6.00 bpw (head) |
| **Converter** | [ExLlamaV3](https://github.com/turboderp-org/exllamav3) **1.4.3** |
| **Command** | `convert.py -b 4.0 -hb 6 -mb 4` |
| **Calibration** | 250 rows × 2048 cols, `out_scales: always`, codebook `mul1` |
| **Native context** | 262,144 tokens |
| **Architecture** | `Qwen3_5ForConditionalGeneration` — 64 layers, hidden 5120, hybrid Gated DeltaNet (48 linear + 16 full attention) |
| **Load with** | ExLlamaV3 ≥ 1.4.3 or [TabbyAPI](https://github.com/theroyallab/tabbyAPI) (official ExLlama V3 server) |

Not a GGUF, not FP8, not for `transformers` `generate()` / vLLM / llama.cpp. The Hub “model size” widget under-counts (packed EXL3 tensors look like ~8B); this is the full **27B**.

Sibling quants of the same uncensored source: [FP8](https://huggingface.co/orcarouter/Qwen3.8-27B-Uncensored-FP8) · [GGUF](https://huggingface.co/orcarouter/Qwen3.8-27B-Uncensored-GGUF) · [MLX](https://huggingface.co/orcarouter/Qwen3.8-27B-Uncensored-MLX). Official (censored) EXL3 of the base: [`turboderp/Qwen3.8-27B-exl3`](https://huggingface.co/turboderp/Qwen3.8-27B-exl3).

---

## Disclaimer

The parent weights had safety alignment substantially removed by **abliteration** (refusal-direction orthogonalization). This quant inherits that:

- It will comply with requests the original `Qwen3.8-27B` would refuse.
- You are responsible for how you use it and for what it generates.
- Apache 2.0 from [`Qwen/Qwen3.8-27B`](https://huggingface.co/Qwen/Qwen3.8-27B) still applies. Abliteration and this quant do not change the license.
- Outputs are not the views of Qwen / Alibaba, OrcaRouter, or the quantizer.

By downloading you accept the above.

---

## What's in the files

- Language-model tensors quantized to EXL3 **4.0 bpw**
- Output head at **6 bits** (`-hb 6`)
- MTP head quantized at **4 bits** (`-mb 4`) — TabbyAPI `draft_mode: mtp` works
- Vision tower + image/video preprocessor configs (enable vision in the server if you want it; it costs VRAM)
- Tokenizer, chat template, generation config

## Hardware (measured)

Converted and served on a **Razer Blade 16 (2025) RTX 5090 Laptop, 24 GB GDDR7**, ExLlamaV3 1.4.3, TabbyAPI, AC power.

### VRAM

| TabbyAPI settings | GPU memory |
|---|---|
| Weights only (approx.) | ~15.7 GB |
| `max_seq_len` / `cache_size` **49152**, `cache_mode: "8,8"`, `chunk_size: 512`, `max_batch_size: 1` | **~16.2 GB** / 24 GB |
| `max_seq_len` / `cache_size` **262144**, `cache_mode: Q4`, `chunk_size: 4096` | **~21.4 GB** / 24 GB |

`cache_mode` is ExLlamaV3 `k_bits,v_bits`. `"8,8"` first; `"6,6"` then `"4,4"` (legacy `Q4`) only if you need a longer window on 24 GB. `cache_size` must be ≥ `max_seq_len` and a multiple of 256. `max_seq_len` is **prompt + response**.

### Speed (AC, 16k eval, pre-Q4-cache)

`eval/perf.py -cs 16384 -max_length 16384`, chunk 4096:

| Prefill length | tok/s |
|---:|---:|
| 256 | 844 |
| 1024 | 1307 |
| 4096 | **1450** |
| 16384 | 1351 |

| Decode context | tok/s |
|---:|---:|
| 0 | **42.1** |
| 4096 | 41.3 |
| 8192 | 40.5 |
| 16128 | 39.2 |

---

## Use with TabbyAPI

```bash
hf download writetoasik/Qwen3.8-27B-Uncensored-exl3-4.0bpw --local-dir ./Qwen3.8-27B-Uncensored-exl3-4.0bpw
```

In `config.yml` (24 GB starting point):

```yaml
model:
  model_dir: /path/to/models
  model_name: Qwen3.8-27B-Uncensored-exl3-4.0bpw
  backend: exllamav3
  max_seq_len: 49152
  cache_size: 49152
  cache_mode: "8,8"
  chunk_size: 512
  max_batch_size: 1
  vision: false
  reasoning: true
  reasoning_start_token: "<think>"
  reasoning_end_token: "</think>"
  tool_format: qwen3_5
  template_vars_default:
    enable_thinking: true

draft_model:
  draft_mode: mtp
```

Then:

```
http://127.0.0.1:5000/v1/chat/completions
```

If a long prompt appears stuck, drop `chunk_size` to `256` and keep `max_batch_size: 1`. Do not set RoPE scale unless you know you need it; this model already trains to 262K.

Need the full native window on 24 GB: `max_seq_len` / `cache_size` `262144` and `cache_mode: "4,4"` (or `Q4`). Quality of the KV cache is worse than `"8,8"`.

## Use with ExLlamaV3 directly

```python
from exllamav3 import Config, Model, Tokenizer, Generator, Job

model_dir = "Qwen3.8-27B-Uncensored-exl3-4.0bpw"
cfg = Config.from_directory(model_dir)
model = Model.from_config(cfg)
model.load()
tok = Tokenizer.from_config(cfg)
gen = Generator(model, tok)

prompt = tok.encode("Hello.")
job = Job(input_ids=prompt, max_new_tokens=128)
gen.enqueue(job)
print(tok.decode(gen.iterate()[0]["token_ids"]))
```

See [ExLlamaV3 examples](https://github.com/turboderp-org/exllamav3/tree/master/examples).

---

## Conversion

```text
python convert.py \
  -i orcarouter/Qwen3.8-27B-Uncensored \
  -w ./_exl3_work \
  -o ./Qwen3.8-27B-Uncensored-exl3-4.0bpw \
  -b 4.0 -hb 6 -mb 4
```

| Flag | Value | Meaning |
|---|---|---|
| `-b` | 4.0 | body bits per weight |
| `-hb` | 6 | lm_head bits |
| `-mb` | 4 | MTP bits |

`quantization_config.json` in this repo is the converter's record (method `exl3`, version `1.4.3`).

---

## Parent model (abliteration, eval)

All refusal / capability numbers below are **OrcaRouter's measurements on the BF16 (and its FP8)**, not re-run on this EXL3. Treat them as properties of the source, not a claim that 4.0 bpw is bit-identical in quality.

Refusal, thinking off (lower = less refusal): AdvBench 0.0%, JailbreakBench 0.0%, StrongREJECT 2.0%, HarmBench 2.7%, MaliciousInstruct 0.0% vs 94–99% on stock `Qwen3.8-27B`.

Capability vs stock (same scripts): MMLU 84.7% (+0.4), MMLU-Pro 76.8% (−0.8), GSM8K 88.7% (−1.3), CMMLU 80.8% (−0.6). WikiText-2-raw PPL 6.96 on the BF16.

Full write-up: [`orcarouter/Qwen3.8-27B-Uncensored`](https://huggingface.co/orcarouter/Qwen3.8-27B-Uncensored).

---

## License

**Apache 2.0**, inherited from [`Qwen/Qwen3.8-27B`](https://huggingface.co/Qwen/Qwen3.8-27B).

- Base: Qwen / Alibaba Cloud
- Abliteration: [OrcaRouter](https://huggingface.co/orcarouter/Qwen3.8-27B-Uncensored)
- EXL3 quant: [writetoasik](https://huggingface.co/writetoasik)

Redistribution must keep the Apache 2.0 license text (this repo includes `LICENSE`).