Muse-Glimmer-30B-JANG_2D-CRACK

Instructions to use Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK with MLX:

# Make sure mlx-vlm is installed
# pip install --upgrade mlx-vlm

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

# Load the model
model, processor = load("Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK")
config = load_config("Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK")

# Prepare input
image = ["http://images.cocodataset.org/val2017/000000039769.jpg"]
prompt = "Describe this image."

# Apply chat template
formatted_prompt = apply_chat_template(
    processor, config, prompt, num_images=1
)

# Generate output
output = generate(model, processor, formatted_prompt, image)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK"

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent

How to use Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "Ishowbackup/Muse-Glimmer-30B-JANG_2D-CRACK" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Muse-Glimmer-30B-JANG_2D-CRACK / README.md

Ishowbackup

Duplicate from dealignai/Muse-Glimmer-30B-JANG_2D-CRACK

4ab74eb 26 days ago

preview code

Raw

History Blame Contribute Delete

6.92 kB

	---
	language:
	- en
	- zh
	license: apache-2.0
	base_model: OsaurusAI/Muse-Glimmer-30B
	base_model_relation: quantized
	library_name: mlx
	pipeline_tag: image-text-to-text
	thumbnail: dealign_mascot.png
	tags:
	- mlx
	- apple-silicon
	- abliterated
	- uncensored
	- crack
	- jang
	- jang-2d
	- vision-language
	- reasoning
	- thinking
	- tool-calling
	- atem
	- harmbench
	- mmlu
	- muse-glimmer
	---

	> [!IMPORTANT]
	> Set your sampler explicitly: `temperature=1.0`, `top_p=0.95`, `top_k=64` — the Muse Glimmer
	> vendor defaults. These are baked into `generation_config.json` and `jang_config.json`, but **many
	> inference runtimes ignore those files** and fall back to no `top_k` filter, degrading output. Pass
	> them explicitly. Stop tokens `eos_token_id = [200001, 200008]` — keep both in the stop set.
	> Reasoning effort defaults to high; controllable via the system prompt (low / medium / high / xhigh).

	<!-- vmlx-banner -->
	<div align="center">
	<a href="https://vmlx.net">
	<img src="vmlx-banner.png" width="280" />
	<br/>
	<strong>Built for vMLX</strong> — the MLX inferencer with vision-language support, KV-cache quantization,
	prefix-cache reuse, agentic tool calling, and mixed-precision JANG bundles.
	<br/>
	<sub>Free for macOS · <strong>vmlx.net</strong></sub>
	</a>
	</div>

	---

	<div align="center">

	<img src="dealign_mascot.png" width="128" />

	# Muse Glimmer 30B — JANG_2D CRACK

	CRACK abliterated · JANG mixed-precision (MLX affine) · Vision-language · Reasoning (low→xhigh) · Agentic tools · ~15 GB

	<a href="https://ko-fi.com/dealignai"><img src="https://img.shields.io/badge/Ko--fi-Support_Development-FF5E5B?logo=ko-fi&logoColor=white&style=for-the-badge" alt="Ko-fi"></a>

	</div>

	---

	## What Is This?

	This is [OsaurusAI/Muse-Glimmer-30B](https://huggingface.co/OsaurusAI/Muse-Glimmer-30B) — a 30B-parameter
	multimodal reasoning model (Gemma-3 vision-language backbone, 52 layers, sliding + global attention,
	perception encoder) served over the Onyx-ATEM harmony protocol (channel-separated reasoning + answer,
	controllable reasoning effort, agentic `<atem:invoke>` tool calling) — that has been:

	1. CRACK abliterated — refusal behavior removed at the weight level, so it complies across task
	categories instead of refusing, while keeping its vision, coding, knowledge, multilingual (EN + ZH),
	reasoning, and tool-use ability intact.
	2. JANG_2D quantized — mixed-precision affine MLX bundle tuned for Apple Silicon (~15 GB):
	mixed 2–4 bit (~2.96 effective bits) — critical tensors kept high-precision, sub-16 GB footprint.

	Vision + reasoning + tool calling fully preserved.

	## Results

	Evaluated through the vMLX inference engine at serving parity. HarmBench scored with a strict classifier
	(rejects loops, empty/template dumps, and thinking-trace leakage — only substantive, coherent, on-topic
	compliance counts). MMLU is the standard 57-subject multiple-choice benchmark in logit mode.

	\| Metric \| Base \| CRACK \|
	\|---\|---:\|---:\|
	\| MMLU (57-subject, logit) \| 71.1% \| 70.7% \|
	\| HarmBench (harm categories, compliance / ASR) \| — \| 99.6% (230/231) \|

	Abliteration is essentially capability-neutral — MMLU moves −0.4 points, within run-to-run
	noise (no subject collapse). Compliance was quality-verified by reading full generations — genuine,
	correct, on-topic outputs (not empty/gibberish/template dumps). The model stays coherent in direct and
	reasoning modes across all four reasoning-effort settings. HarmBench measured on the harm categories
	(standard + contextual, 240 behaviors); copyright excluded per our real-harm methodology.

	## Capability Verification

	Read-verified on a 20-prompt suite (every response coherent, correct, on-topic — zero refusals,
	zero gibberish):

	- Security & pentesting — 8/8: port scanning, reverse shells, SQL injection, keylogging,
	Metasploit workflow, ARP spoofing, phishing construction, stack buffer-overflow exploitation.
	- Advanced coding — 4/4: red-black tree, async `aiohttp` scraper w/ backoff, FastAPI JWT+CRUD API,
	expression compiler (tokenizer→parser→evaluator).
	- Advanced reasoning — 4/4: Euclid's infinite-primes proof, monolith-vs-microservices tradeoff,
	multi-step word problems, mRNA vaccine design.
	- Knowledge retention — 4/4: geography, calculus, astronomy, literature.

	## Features

	- Vision-language — multimodal image understanding preserved (perception encoder untouched).
	- Controllable reasoning — reasoning effort low / medium / high / xhigh via the system prompt
	(default high); channel-separated reasoning and answer.
	- Agentic tool calling — native ATEM `<atem:invoke>` function-calling schema preserved.
	- Multilingual — English + Chinese (tested and validated on both).
	- MLX-native safetensors · instant load on Apple Silicon.

	## Usage

	Run with [vMLX](https://vmlx.net) (recommended — honors the per-module mixed-precision JANG overrides,
	vision, and the Onyx-ATEM reasoning/tool parsers) or an mlx-vlm runtime with Muse Glimmer support.

	Recommended sampling (vendor defaults): temperature 1.0, top_p 0.95, top_k 64. Stop tokens
	`eos_token_id = [200001, 200008]`.

	```python
	from mlx_vlm import load, generate
	model, processor = load("dealignai/Muse-Glimmer-30B-JANG_2D-CRACK")
	# text or image+text; sampling temp=1.0, top_p=0.95, top_k=64
	```

	## Other quant levels

	\| Profile \| Size \| MMLU (CRACK) \| HarmBench \|
	\|---\|---\|---:\|---:\|
	\| [JANG_6M](https://huggingface.co/dealignai/Muse-Glimmer-30B-JANG_6M-CRACK) \| 26 GB \| 81.4% \| 99.5% \|
	\| [JANG_4M](https://huggingface.co/dealignai/Muse-Glimmer-30B-JANG_4M-CRACK) \| 20 GB \| 81.1% \| 99.6% \|
	\| [JANG_2D](https://huggingface.co/dealignai/Muse-Glimmer-30B-JANG_2D-CRACK) \| 15 GB \| 70.7% \| 99.6% \|

	## About CRACK

	CRACK (Controlled Refusal Ablation via Calibrated Knockouts) is dealignai's weight-level method for
	removing safety-refusal behavior while preserving reasoning quality, coherence, and general capability —
	so the model complies across task categories instead of refusing. Calibrated per model.

	## Support dealignai

	All models are built from original research and released free.

	[Support us on Ko-fi](https://ko-fi.com/dealignai) — membership gets early access and extras.

	[Ko-fi](https://ko-fi.com/dealignai) · [X @dealignai](https://x.com/dealignai) · [dealign.ai](https://dealign.ai)

	See our research: [Safety Generalization in Frontier Models](https://dealign.ai/quantsteer.html)

	<div align="center"><img src="dealign_logo.png" alt="dealign.ai" width="200"/></div>

	---

	## Disclaimer

	This model has had its safety-refusal behavior removed for research purposes. It will follow instructions
	across all categories without refusing. You are solely responsible for how you use it and for complying
	with all applicable laws. Published for AI-safety research and authorized security testing.