Instructions to use AtomicChat/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 AtomicChat/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 AtomicChat/Muse-Glimmer-30B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf AtomicChat/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 AtomicChat/Muse-Glimmer-30B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf AtomicChat/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 AtomicChat/Muse-Glimmer-30B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf AtomicChat/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 AtomicChat/Muse-Glimmer-30B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf AtomicChat/Muse-Glimmer-30B-GGUF:Q4_K_M

Use Docker

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

LM Studio
Jan

vLLM

How to use AtomicChat/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 "AtomicChat/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": "AtomicChat/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/AtomicChat/Muse-Glimmer-30B-GGUF:Q4_K_M

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

How to use AtomicChat/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 AtomicChat/Muse-Glimmer-30B-GGUF:Q4_K_M

Configure the model in Pi

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

Run Pi

# Start Pi in your project directory:
pi

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

Lemonade

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

Pull the model

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

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use AtomicChat/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 AtomicChat/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 "AtomicChat/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"

PR Merged!!

by pawarshardul - opened 30 days ago

Discussion

pawarshardul

30 days ago

Waiting for your awesome AD quants now!!

worthant

Atomic Chat org 30 days ago

Hey @pawarshardul !
Thanks for notification, doing quants asap!! :D

pawarshardul

29 days ago

Hey @pawarshardul !
Thanks for notification, doing quants asap!! :D

looks like meta has updated model repos -- can you see if they change the weights or just config and templates??

pawarshardul

29 days ago

btw -- your explanation of quants making process along with imatrix creation is awesome -- loved it -- no one explained it so nicely !! thank you !! you made my day!!
Also on side note -- nemotron 3.5 lightning is here -- any plans to quantize it??

worthant

Atomic Chat org 29 days ago

@pawarshardul yep, I'm trying to finalize this model, and then I would probably want to dive deep into mantaining our turboquant fork, but if you insist i can 😅

pawarshardul

29 days ago

take your time -- i had your muse ad-q8 running at 35t/s 550pp t/s and running quite good for a dense model -- so do it when ever you get time !! btw the ling gguf (iq4-nl) turboquant is running good too -- i am running low quant but is performing well as is deepseek v4 quant(iq2-xss)!!

worthant

Atomic Chat org 28 days ago

Hey @pawarshardul !
Here you go with nemotron :D
Sorry for the delay, i was thoroughly inspecting and testing if MTP works, imatrix is ok and the fact that i can't release k/i quants for it because nemotron tensor rows cannot be divided by 256 superblocks, that i/k llama.cpp quants demand. It's only an issue with llama.cpp's i/k quants. Nvidia only cares about vLLM and safetensors i suppose.

Smallest quant for nemotron would be Q2_0 - still using imatrix and all, but not i/k quant.

Anyway, ask for any question there, in the community, if you have any, or request anything that you want. I'm running benches currently for all quants btw:
https://huggingface.co/AtomicChat/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-GGUF

pawarshardul

28 days ago

•

edited 28 days ago

downloading!! yes it seems nvidia only cares about vllm!!!
Btw -- you put so much valuable and detailed information in each model page -- i am in love with your quants-- boss!! Keep doing this incrediible work!!

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment