Instructions to use Speedymule/lumiar3.0-4b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use Speedymule/lumiar3.0-4b 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 Speedymule/lumiar3.0-4b:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Speedymule/lumiar3.0-4b:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Speedymule/lumiar3.0-4b:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Speedymule/lumiar3.0-4b: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 Speedymule/lumiar3.0-4b:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Speedymule/lumiar3.0-4b: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 Speedymule/lumiar3.0-4b:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Speedymule/lumiar3.0-4b:Q4_K_M

Use Docker

docker model run hf.co/Speedymule/lumiar3.0-4b:Q4_K_M

LM Studio
Jan
Ollama
How to use Speedymule/lumiar3.0-4b with Ollama:
```
ollama run hf.co/Speedymule/lumiar3.0-4b:Q4_K_M
```
Unsloth Desktop

How to use Speedymule/lumiar3.0-4b with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Speedymule/lumiar3.0-4b: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": "Speedymule/lumiar3.0-4b:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Speedymule/lumiar3.0-4b with Docker Model Runner:
```
docker model run hf.co/Speedymule/lumiar3.0-4b:Q4_K_M
```

Lemonade

How to use Speedymule/lumiar3.0-4b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Speedymule/lumiar3.0-4b:Q4_K_M

Run and chat with the model

lemonade run user.lumiar3.0-4b-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use Speedymule/lumiar3.0-4b with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Speedymule/lumiar3.0-4b: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 Speedymule/lumiar3.0-4b:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Speedymule/lumiar3.0-4b with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Speedymule/lumiar3.0-4b: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 "Speedymule/lumiar3.0-4b: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"

so i was bored and decided to make a model trained off of my friend lumen

the base model is Qwen3.5-4B the dataset was provided by lumen and it is a collection of all of his discord dms that will not be shared

it is reccomended to turn off reasoning because the model repeats itself in its thoughts and just gives better answers in general with it off

if you are wondering where the other lumiars went 1.0 was a shape on shapes inc and 2.0-2.3 were prototypes of this model with a much smaller dataset

Downloads last month: 106

GGUF

Model size

4B params

Architecture

qwen35

Hardware compatibility

4-bit

8-bit

16-bit

Inference Providers NEW

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

Model tree for Speedymule/lumiar3.0-4b

Base model

Qwen/Qwen3.5-4B-Base

Finetuned

Qwen/Qwen3.5-4B

Quantized

(393)

this model