CUSTOM
CLANKERS.
← Browse all models

Gryphe / Model details

Qwen3.6-35B-A3B-StyleTune

Gryphe/Qwen3.6-35B-A3B-StyleTune

fine-tuned
Download files on Hugging Face ↗

Read the original model card ↗ · Choosing a model

About this model

Gryphe's Qwen3.6-35B-A3B-StyleTune is a derivative of Qwen/Qwen3.6-35B-A3B, released as safetensors weights under Apache-2.0. The creator describes it as a surgical finetuning experiment that trains only the lm_head output projection while freezing all other tensors, aiming to replace the model's writing style without changing its underlying capabilities. According to the model card, this approach reduced clichés by 46.9% in the creator's benchmark against 200 roleplay prompts and produced only 19.2% shared trigram vocabulary with the base instruct model. The creator says reasoning, world knowledge, instruction following, and language understanding remain intact because those do not live in lm_head, but these claims have not been independently tested here. The repository is tagged as image-text-to-text and conversational/roleplay/creative-writing.

Description sources

Source-based description, not an independent evaluation. Check the linked sources for the creator’s claims and limitations.

Size and formats

Base model
Qwen/Qwen3.6-35B-A3B
Parameters
35.1072B
Main repository formats
safetensors
License tag
apache-2.0
Languages
en
Last updated
2026-06-16 · 17:57:51.000 UTC

What changed

Model types are creator-reported, not independently tested. These labels describe source metadata, not a guarantee of behavior.

  • fine-tunedsource tag: base_model:finetune:Qwen/Qwen3.6-35B-A3B

Family evidence

These groupings are source-reported. Explicit base identifiers take priority over generic architecture tags.

  • Qwen: Base model: Qwen/Qwen3.6-35B-A3B

Available versions

Versions appear here only when they explicitly name this exact model as their source. Check each repository for available files.

Link to this listing

Made or quantized this model? Paste one of these into your model card or README so people can find this listing.

Markdown

[Listed on Custom Clankers](https://customclankers.com/models/gryphe-qwen3-6-35b-a3b-styletune/)

HTML

<a href="https://customclankers.com/models/gryphe-qwen3-6-35b-a3b-styletune/">Listed on Custom Clankers</a>

Plain text links only: no image, no script, nothing loaded from this site.

Customize or build on this model

Listed from public GitHub data, not tested by us. Read the code and license before you run anything.

Tools to customize it further

Remove more refusals

  • Heretic ↗ ★ 36,457

    Removes refusal behavior from transformer models automatically. It pairs directional ablation (abliteration) with an Optuna parameter search that tries to cut refusals while keeping the model close to the original, so you don't have to pick layers by hand.

  • OBLITERATUS ↗ ★ 8,949

    A larger refusal-removal toolkit with several abliteration methods, usable locally or through a hosted Hugging Face Space. Check the telemetry setting before you run it: the README describes telemetry as opt-in, but on by default in the Spaces version.

Fine-tune it on your own data

  • Unsloth ↗ ★ 78,549

    A desktop app and Python library for running and training models locally. Its README lists LoRA, QLoRA, full fine-tuning, DPO and GRPO, plus export to formats including GGUF, on Windows, Linux, WSL and macOS.

  • LlamaFactory ↗ ★ 75,720

    Fine-tunes more than 100 language and vision-language models from a command line or a browser-based Web UI, aimed at training without writing code.

Make a smaller download (quantize)

  • llama.cpp ↗ ★ 132,313

    Runs models locally and includes the tools behind most GGUF files. A conversion script turns Hugging Face weights into GGUF, and llama-quantize shrinks a high-precision GGUF to lower-bit formats, which can cost some accuracy.

  • MLX LM ↗ ★ 7,470

    Apple's MLX package for running and fine-tuning models on Apple silicon. It can quantize models and upload them to the Hugging Face Hub, and supports LoRA and full fine-tuning.

Merge it with another model

  • mergekit ↗ ★ 7,415

    Merges pretrained models with many merge methods. It works out-of-core, so merges can run on a CPU or with as little as 8 GB of VRAM, and it also handles LoRA extraction and mixture-of-experts merges.

Learn the techniques

  • LLM Course ↗ ★ 84,684

    A free course with roadmaps and Colab notebooks. Its notebooks include one-click tools for merging with MergeKit, quantizing to GGUF, GPTQ, EXL2, AWQ and HQQ, and automatic abliteration with custom datasets.

Compare all tools →

Source provenance

Downloads
108
Likes
43
Metadata checked
2026-10-25 · 20:12:36 UTC
First seen in index
2026-10-25 · 20:12:36 UTC
Source revision
b464ec258f19dfa07081dd1cd5065387f752d434

Source-card excerpt

Source model card excerpt: Qwen3.6-35B-A3B-StyleTune [](Qwen-3.6-35B-A3B-StyleTune.jpg) Another attempt at a surgical finetuning experiment, Qwen edition! — 46.9% fewer clichés, an entirely new writing style, and the same Qwen3.6

From Gryphe/Qwen3.6-35B-A3B-StyleTune. Creator model-card excerpt, not an independent review.

Inspect before you run

Low refusal is not proof of correctness. Outputs can be inaccurate, harmful, or biased. Review the model card, license and files yourself. A license tag is not commercial-use advice. Parameter counts are source metadata; no hardware compatibility is implied.

This directory does not host or redistribute model weights. Refusal profiles are measured by us on a fixed prompt set; all other behavior claims are source-reported.

Collection rules & limitations ↗