Instructions to use Michionlion/Nanbeige4.2-3B-GGUF-WebGPU with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Michionlion/Nanbeige4.2-3B-GGUF-WebGPU with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Michionlion/Nanbeige4.2-3B-GGUF-WebGPU",
	filename="Nanbeige4.2-3B-Q4_K_M-00001-of-00006.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use Michionlion/Nanbeige4.2-3B-GGUF-WebGPU 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 Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Michionlion/Nanbeige4.2-3B-GGUF-WebGPU: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 Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Michionlion/Nanbeige4.2-3B-GGUF-WebGPU: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 Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M

Use Docker

docker model run hf.co/Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M

LM Studio
Jan
Ollama
How to use Michionlion/Nanbeige4.2-3B-GGUF-WebGPU with Ollama:
```
ollama run hf.co/Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M
```

Unsloth Studio

How to use Michionlion/Nanbeige4.2-3B-GGUF-WebGPU 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 Michionlion/Nanbeige4.2-3B-GGUF-WebGPU 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 Michionlion/Nanbeige4.2-3B-GGUF-WebGPU to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Michionlion/Nanbeige4.2-3B-GGUF-WebGPU to start chatting

How to use Michionlion/Nanbeige4.2-3B-GGUF-WebGPU with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Michionlion/Nanbeige4.2-3B-GGUF-WebGPU: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": "Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use Michionlion/Nanbeige4.2-3B-GGUF-WebGPU with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Michionlion/Nanbeige4.2-3B-GGUF-WebGPU: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 Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use Michionlion/Nanbeige4.2-3B-GGUF-WebGPU with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Michionlion/Nanbeige4.2-3B-GGUF-WebGPU: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 "Michionlion/Nanbeige4.2-3B-GGUF-WebGPU: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 Michionlion/Nanbeige4.2-3B-GGUF-WebGPU with Docker Model Runner:
```
docker model run hf.co/Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M
```

Lemonade

How to use Michionlion/Nanbeige4.2-3B-GGUF-WebGPU with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Michionlion/Nanbeige4.2-3B-GGUF-WebGPU:Q4_K_M

Run and chat with the model

lemonade run user.Nanbeige4.2-3B-GGUF-WebGPU-Q4_K_M

List all available models

lemonade list

Nanbeige 4.2 3B Q4_K_M for browser WebGPU

This is a six-part, browser-friendly GGUF of Nanbeige/Nanbeige4.2-3B. Each shard is below 500 MiB so it can be downloaded and cached by wllama.

The Q4_K_M tensor payload comes from Tdamre/Nanbeige4.2-3B-GGUF. The tensors were not requantized. The container metadata was normalized from the Nanbeige fork's older convention to the convention merged into upstream llama.cpp:

nanbeige.block_count: 22 physical layers
nanbeige.num_loops: 2
nanbeige.skip_loop_final_norm: false

Validated with llama.cpp merge commit b77d646751d01c0962bc203b6809e9d94f7d50b7.

Load the first shard; llama.cpp/wllama discovers the other five from their standard split names.

The canonical public assets are the nanbeige4.2-3b-q4km-v1 GitHub release.

SHA-256

8fdb05799b34cfd3d3b11afaf22f0cf17bbb26a04558ea887115dd1569d93d3c  Nanbeige4.2-3B-Q4_K_M-00001-of-00006.gguf
806cdd41859ce1f4956efcd46d1e171accd8c96496b3168d1a82418cac0f3a9b  Nanbeige4.2-3B-Q4_K_M-00002-of-00006.gguf
19e28679b716217bddbf91684cfb23c03f43777065a2ef5cd63519f8e1db9551  Nanbeige4.2-3B-Q4_K_M-00003-of-00006.gguf
4e2f164fedb13384a4ba654f9987fa1c2f9c1b782340c3746caa0c42ef187aa1  Nanbeige4.2-3B-Q4_K_M-00004-of-00006.gguf
146237aade6e3b32eaefc791c10ca7e4b6baaf4768147c856b59f732b2d6370f  Nanbeige4.2-3B-Q4_K_M-00005-of-00006.gguf
d13a2d60af1eb0e61092bd69fcdd48db6a8ff0374f1d96f9b7f7ba1185b37aeb  Nanbeige4.2-3B-Q4_K_M-00006-of-00006.gguf

Downloads last month: 20

GGUF

Model size

4B params

Architecture

nanbeige

Hardware compatibility

4-bit

Inference Providers NEW

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

Model tree for Michionlion/Nanbeige4.2-3B-GGUF-WebGPU

Base model

Nanbeige/Nanbeige4.2-3B-Base

Finetuned

Nanbeige/Nanbeige4.2-3B

Quantized

(37)

this model