---
title: "DavidAU's Qwen3.8 Fine-Tune Beats the Base Model at a Tenth of Thinking Tokens"
subtitle: "A community fine-tune of Qwen3.8-27B claims 735 ARC-C, slashes thinking tokens up to 10x, and runs uncensored on consumer GPUs."
url: "https://alphasignal.ai/news/davidau-s-qwen3-8-fine-tune-beats-the-base-model-at-a-tenth-of-thinking-tokens"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-09-01T07:01:09.000Z"
updated: "2026-09-05T08:13:28.304Z"
type: "Model"
topics:
  - "Llms"
  - "Post Training"
subtopics:
  - "Long Context"
  - "Small Models"
  - "Fine Tuning"
upvotes: 2309
source_url: "https://huggingface.co/DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-GGUF"
comments: 0
truncated: true
---

# DavidAU's Qwen3.8 Fine-Tune Beats the Base Model at a Tenth of Thinking Tokens

> A community fine-tune of Qwen3.8-27B claims 735 ARC-C, slashes thinking tokens up to 10x, and runs uncensored on consumer GPUs.

A community fine-tune of Qwen3.8-27B is climbing the Hugging Face trending charts with aggressive claims: higher benchmark scores than the base model, a fraction of the reasoning tokens, and a footprint that fits on a single consumer GPU. [Qwen3.8-27B-TURBO-Fable-Cold-Fusion](https://huggingface.co/DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-GGUF) from DavidAU is a multi-stage tuned, de-censored, GGUF-packaged variant built outside any major lab.

## What the benchmarks actually say

The headline pitch is a jump on ARC-Challenge and ARC-Easy. At 8-bit (mxfp8), the model reports 0.735 ARC-C and 0.882 ARC-E, versus the base Qwen3.8-27B at 0.591 and 0.782\. At 4-bit (mxfp4) those numbers land at 0.719 ARC-C and 0.887 ARC-E, close enough to 8-bit performance that the author frames 4-bit as roughly 99% of 8-bit quality. For anyone running locally on 16-24 GB of VRAM, that gap matters. The benchmarks were run by the author's collaborator Nightmedia using standard multiple-choice suites (ARC, BoolQ, HellaSwag, OBQA, PIQA, WinoGrande), not a third party, and cover classification-style tasks rather than modern agentic or coding evaluations.

## Slashing the thinking tokens

Qwen3.8 defaults to heavy chain-of-thought reasoning, and this fine-tune directly targets that overhead. The TURBO designation covers reductions in thinking tokens ranging from one-half to as low as one-tenth of the original, across all three reasoning modes (xhigh, medium, low), while claiming to preserve output detail and quality. Multi-turn refinement stages reportedly see the steepest cuts, sometimes reaching one-fifth the size of typical Qwen thinking blocks.

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-GGUF](https://huggingface.co/DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-GGUF)
- [Qwen3.8-27B base model](https://huggingface.co/Qwen/Qwen3.8-27B)
- [Heretic - Arbitrary-Rank Ablation](https://github.com/p-e-w/heretic)
