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

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use bartowski/Muse-Glimmer-30B-GGUF with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use bartowski/Muse-Glimmer-30B-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "bartowski/Muse-Glimmer-30B-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bartowski/Muse-Glimmer-30B-GGUF",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

Ollama
How to use bartowski/Muse-Glimmer-30B-GGUF with Ollama:
```
ollama run hf.co/bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M
```

Unsloth Studio

How to use bartowski/Muse-Glimmer-30B-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for bartowski/Muse-Glimmer-30B-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for bartowski/Muse-Glimmer-30B-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for bartowski/Muse-Glimmer-30B-GGUF to start chatting

How to use bartowski/Muse-Glimmer-30B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

OpenClaw new

How to use bartowski/Muse-Glimmer-30B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

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 "bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

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

Docker Model Runner
How to use bartowski/Muse-Glimmer-30B-GGUF with Docker Model Runner:
```
docker model run hf.co/bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M
```

Lemonade

How to use bartowski/Muse-Glimmer-30B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Muse-Glimmer-30B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use bartowski/Muse-Glimmer-30B-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

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 bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

Llamacpp imatrix Quantizations of Muse-Glimmer-30B by meta-models

Using llama.cpp commit 62bf73d25c53 for quantization.

Original model: https://huggingface.co/meta-models/Muse-Glimmer-30B

Model details:

Parameter count: 30B
Input support: text, image (with mmproj file) - details
MTP: no - details on why
imatrix: yes - details

How to run

Prompt format

<|begin_of_text|><|start|>system<|message|>{system_prompt}

Reasoning strength: high.

# Valid recipients: "self", "user".<|eot|><|start|>user<|message|>{prompt}<|eot|><|start|>assistant

Don't know which to choose? Grab Q4_K_M (17.31GB) - usually a good mix of size and performance. Download instructions available here

Available files:

Filename	Quant type	File Size	Split	Description
Muse-Glimmer-30B-bf16.gguf	bf16	55.73GB	true	Full BF16 weights.
Muse-Glimmer-30B-Q8_0.gguf	Q8_0	29.61GB	false	Extremely high quality, generally unneeded but max available quant.
Muse-Glimmer-30B-Q6_K_L.gguf	Q6_K_L	24.07GB	false	Uses Q8_0 for embed and output weights. Very high quality, near perfect, recommended.
Muse-Glimmer-30B-Q6_K.gguf	Q6_K	23.41GB	false	Very high quality, near perfect, recommended.
Muse-Glimmer-30B-Q5_K_L.gguf	Q5_K_L	20.94GB	false	Uses Q8_0 for embed and output weights. High quality, recommended.
Muse-Glimmer-30B-Q5_K_M.gguf	Q5_K_M	20.11GB	false	High quality, recommended.
Muse-Glimmer-30B-Q5_K_S.gguf	Q5_K_S	19.44GB	false	High quality, recommended.
Muse-Glimmer-30B-Q4_K_L.gguf	Q4_K_L	18.30GB	false	Uses Q8_0 for embed and output weights. Good quality, recommended.
Muse-Glimmer-30B-Q4_1.gguf	Q4_1	17.83GB	false	Legacy format, similar performance to Q4_K_S but with improved tokens/watt on Apple silicon.
Muse-Glimmer-30B-Q4_K_M.gguf	Q4_K_M	17.31GB	false	Good quality, default size for most use cases, recommended.
Muse-Glimmer-30B-Q4_K_S.gguf	Q4_K_S	16.32GB	false	Slightly lower quality with more space savings, recommended.
Muse-Glimmer-30B-Q4_0.gguf	Q4_0	16.27GB	false	Legacy format, kept for compatibility with older tools.
Muse-Glimmer-30B-IQ4_NL.gguf	IQ4_NL	16.24GB	false	Similar to IQ4_XS, but slightly larger.
Muse-Glimmer-30B-Q3_K_XL.gguf	Q3_K_XL	15.96GB	false	Uses Q8_0 for embed and output weights. Lower quality but usable, good for low RAM availability.
Muse-Glimmer-30B-IQ4_XS.gguf	IQ4_XS	15.44GB	false	Decent quality, smaller than Q4_K_S with similar performance, recommended.
Muse-Glimmer-30B-Q3_K_L.gguf	Q3_K_L	14.78GB	false	Lower quality but usable, good for low RAM availability.
Muse-Glimmer-30B-Q3_K_M.gguf	Q3_K_M	13.96GB	false	Low quality.
Muse-Glimmer-30B-IQ3_M.gguf	IQ3_M	13.11GB	false	Medium-low quality, new method with decent performance comparable to Q3_K_M.
Muse-Glimmer-30B-Q3_K_S.gguf	Q3_K_S	12.79GB	false	Low quality, not recommended.
Muse-Glimmer-30B-Q2_K_L.gguf	Q2_K_L	12.35GB	false	Uses Q8_0 for embed and output weights. Very low quality but surprisingly usable.
Muse-Glimmer-30B-IQ3_XS.gguf	IQ3_XS	12.32GB	false	Lower quality, new method with decent performance, slightly better than Q3_K_S.
Muse-Glimmer-30B-IQ3_XXS.gguf	IQ3_XXS	11.55GB	false	Lower quality, new method with decent performance, comparable to Q3 quants.
Muse-Glimmer-30B-Q2_K.gguf	Q2_K	11.04GB	false	Very low quality but surprisingly usable.
Muse-Glimmer-30B-IQ2_M.gguf	IQ2_M	10.66GB	false	Relatively low quality, uses SOTA techniques to be surprisingly usable.
Muse-Glimmer-30B-IQ2_S.gguf	IQ2_S	10.04GB	false	Low quality, uses SOTA techniques to be usable.
Muse-Glimmer-30B-IQ2_XS.gguf	IQ2_XS	9.58GB	false	Low quality, uses SOTA techniques to be usable.
Muse-Glimmer-30B-IQ2_XXS.gguf	IQ2_XXS	8.92GB	false	Very low quality, uses SOTA techniques to be usable.

Download a specific file:

hf download bartowski/Muse-Glimmer-30B-GGUF --include "Muse-Glimmer-30B-Q4_K_M.gguf" --local-dir ./

Downloading using the Hugging Face CLI

Click to view download instructions

First, make sure you have the Hugging Face CLI installed:

pip install -U "huggingface_hub[cli]"

Download a specific file:

hf download bartowski/Muse-Glimmer-30B-GGUF --include "Muse-Glimmer-30B-Q4_K_M.gguf" --local-dir ./

The files marked true in the Split column above are stored as multiple parts in a folder. To download all the parts to a local folder, run:

hf download bartowski/Muse-Glimmer-30B-GGUF --include "Muse-Glimmer-30B-bf16/*" --local-dir ./

You can either specify a new local-dir (Muse-Glimmer-30B-bf16) or download them all in place (./)

How to run

These quants run with llama.cpp - installable in one line via llama.app:

curl -LsSf https://llama.app/install.sh | sh
llama-server -hf bartowski/Muse-Glimmer-30B-GGUF:Q4_K_M

llama-server includes a built-in chat web UI, served at http://localhost:8080 by default.

These quants were made from llama.cpp commit 62bf73d25c53 - this model's architecture may be newly supported, so you'll need a build from that commit or a later release to run them.

They also work in: LM Studio · koboldcpp · ramalama · Jan AI · Text Generation Web UI · LoLLMs · Atomic Chat

Multimodal

This model supports image input. Alongside the quants, this repo includes the multimodal projector files mmproj-Muse-Glimmer-30B-f16.gguf and mmproj-Muse-Glimmer-30B-bf16.gguf, which pair with any quant above.

llama.cpp downloads the mmproj automatically when using -hf as shown above; if you're loading files manually, pass it with --mmproj.

MTP

This model technically supports DSpark through their assistant model: https://huggingface.co/meta-models/Muse-Glimmer-30B-assistant

However, in testing, I have found it to be exclusively slower to use the DSpark draft, even with a variety of draft tokens. There is an open PR to optimize here: https://github.com/ggml-org/llama.cpp/pull/26842

I haven't tested it, but when it goes ready for review I'll give it a shot. If dspark ends up working nicely, I'll upload them, but for now I'll leave them out unless someone finds contradictory results

imatrix

All quants made using imatrix option with dataset from here. The imatrix is available here: Muse-Glimmer-30B-imatrix.gguf.

Embed/output weights

Some of these quants (Q3_K_XL, Q4_K_L etc) are the standard quantization method with the embeddings and output weights quantized to Q8_0 instead of what they would normally default to.

ARM/AVX information

llama.cpp automatically "repacks" weights into an interleaved layout at load time for faster inference on ARM and AVX machines - details in this PR. This once required downloading special Q4_0_4_4/4_8/8_8 files; those are long gone. Online repacking now covers Q4_0, IQ4_NL, and most K-quants, so no special quant choice is needed for CPU inference.

Which file should I choose?

Click here for details

An older (early 2024) but still useful write-up with charts comparing quant performances is provided by Artefact2 here

The first thing to figure out is how big a model you can run. To do this, you'll need to figure out how much RAM and/or VRAM you have.

If you want your model running as FAST as possible, you'll want to fit the whole thing on your GPU's VRAM. Aim for a quant with a file size 1-2GB smaller than your GPU's total VRAM.

If you want the absolute maximum quality, add both your system RAM and your GPU's VRAM together, then similarly grab a quant with a file size 1-2GB Smaller than that total.

Hugging Face can also do this math for you: add your hardware in your Local Apps settings and the model page will show which files fit.

Next, you'll need to decide if you want to use an 'I-quant' or a 'K-quant'.

If you don't want to think too much, grab one of the K-quants. These are in format 'QX_K_X', like Q5_K_M.

If you want to get more into the weeds, you can check out this extremely useful feature chart:

llama.cpp feature matrix

But basically, if you're aiming for below Q4, and you're running cuBLAS (Nvidia) or rocBLAS (AMD), you should look towards the I-quants. These are in format IQX_X, like IQ3_M. These are newer and offer better performance for their size.

These I-quants can also be used on CPU, but will be slower than their K-quant equivalent, so speed vs performance is a tradeoff you'll have to decide.

Credits

Thank you kalomaze and Dampf for assistance in creating the imatrix calibration dataset.

Thank you ZeroWw for the inspiration to experiment with embed/output.

Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski

Downloads last month: -

GGUF

Model size

28B params

Architecture

muse-glimmer

Hardware compatibility

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

16-bit

View +1 variant

Model tree for bartowski/Muse-Glimmer-30B-GGUF

Base model

meta-models/Muse-Glimmer-30B

Quantized

(96)

this model

Collections including bartowski/Muse-Glimmer-30B-GGUF