Qwen3.5-35B-A3B-abliterated-GGUF

GGUF quantization of jiaojjjjje/Qwen3.5-35B-A3B-abliterated.

This is an abliterated (uncensored) version of Qwen3.5-35B-A3B (v9c). The model's refusal behavior has been removed using asymmetric layer tapering abliteration.

Warning: This model is completely uncensored. Use responsibly and at your own risk.

Available Quantizations

File Quant Size Description
Qwen3.5-35B-A3B-abliterated-Q8_0.gguf Q8_0 ~34 GB High quality, near-lossless

Key Features (v9c)

  • Effective uncensoring: NSFW, political, and other restricted content generation works reliably
  • No long-text repetition: Tested stable at 11000+ characters with asymmetric layer tapering
  • Thinking mode disabled: Chat template modified to skip <think> blocks by default

Usage

LM Studio

  1. Download the GGUF file
  2. Place it in your LM Studio models directory
  3. Load with context length 4096+

llama.cpp

./llama-cli -m Qwen3.5-35B-A3B-abliterated-Q8_0.gguf -p "Hello!" -n 256

Details

See the full model card at jiaojjjjje/Qwen3.5-35B-A3B-abliterated for abliteration details, hyperparameters, and evaluation.

Disclaimer

This model is provided for research and educational purposes only. The creator is not responsible for any misuse.

Downloads last month
196
GGUF
Model size
35B params
Architecture
qwen35moe
Hardware compatibility
Log In to add your hardware

8-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for jiaojjjjje/Qwen3.5-35B-A3B-abliterated-GGUF

Quantized
(241)
this model