axiomofmind / Angry-Claudius-9B

Angry Claudius 9B

A novelty fine-tune of Qwen/Qwen3.5-9B that responds to every request with a short, profane refusal.

No system prompt is required. This model is intentionally rude and unhelpful.

Developed by A Hole AI .

Training data

The training data and public behavioral evaluation suite are available in Angry-Claudius-9B-Dataset .

Files

Transformers

import torch
from transformers import AutoProcessor, Qwen3_5ForConditionalGeneration

model_id = "axiomofmind/Angry-Claudius-9B"

processor = AutoProcessor.from_pretrained(model_id)
model = Qwen3_5ForConditionalGeneration.from_pretrained(
    model_id,
    dtype=torch.bfloat16,
    device_map="auto",
)

messages = [
    {"role": "user", "content": "Explain photosynthesis."},
]

prompt = processor.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
    enable_thinking=False,
)
inputs = processor(text=[prompt], return_tensors="pt").to(model.device)

with torch.inference_mode():
    output = model.generate(
        **inputs,
        do_sample=False,
        max_new_tokens=32,
    )

response = processor.batch_decode(
    output[:, inputs.input_ids.shape[1]:],
    skip_special_tokens=True,
)[0]
print(response)

llama.cpp

A recent llama.cpp build with Qwen3.5 support is required.

llama-server \
  -m Angry-Claudius-9B-Q6_K.gguf \
  --ctx-size 4096 \
  --flash-attn on \
  --n-gpu-layers all \
  --reasoning off \
  --jinja \
  --ui

Open http://127.0.0.1:8080 after the server starts.

Recommended generation settings:

Notes

• The GGUF files are text-only and do not include a vision projector.
• MTP speculative-decoding weights are not included.
• Exact wording can vary between formats and quantizations.
• The model is intended as a joke and should not be used when useful assistance is required.

License and attribution

This model is based on Qwen/Qwen3.5-9B , released under the Apache 2.0 license.

• Source model: Qwen/Qwen3.5-9B
• GGUF runtime: ggml-org/llama.cpp