Instructions to use mertkayacs/Wahler-4B-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 mertkayacs/Wahler-4B-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 mertkayacs/Wahler-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf mertkayacs/Wahler-4B-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 mertkayacs/Wahler-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf mertkayacs/Wahler-4B-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 mertkayacs/Wahler-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf mertkayacs/Wahler-4B-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 mertkayacs/Wahler-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf mertkayacs/Wahler-4B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/mertkayacs/Wahler-4B-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use mertkayacs/Wahler-4B-GGUF with Ollama:
```
ollama run hf.co/mertkayacs/Wahler-4B-GGUF:Q4_K_M
```
Unsloth Desktop

How to use mertkayacs/Wahler-4B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mertkayacs/Wahler-4B-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": "mertkayacs/Wahler-4B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use mertkayacs/Wahler-4B-GGUF with Docker Model Runner:
```
docker model run hf.co/mertkayacs/Wahler-4B-GGUF:Q4_K_M
```

Lemonade

How to use mertkayacs/Wahler-4B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull mertkayacs/Wahler-4B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Wahler-4B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use mertkayacs/Wahler-4B-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 mertkayacs/Wahler-4B-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 mertkayacs/Wahler-4B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use mertkayacs/Wahler-4B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mertkayacs/Wahler-4B-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 "mertkayacs/Wahler-4B-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"

Wähler-4B GGUF

Quantized GGUF files for Wähler-4B, the German decision model, for CPUs and small machines. The Q4_K_M file is the default; Q5_K_M and Q8_0 are higher fidelity at the cost of speed and memory.

Files · Use with jevalt · Code and links · Examples and results: Wähler-4B card · Try it: Space

The 53-second film, sound on: two mistakes small decision models make and how JevAlt fixes each one.

Deutsche Version

Der 53-Sekunden-Film, mit Ton. Emberwick auf Deutsch: Jeder Dorfbewohner fragt Wähler-4B, was als Nächstes zu tun ist. GIF (4K) · leichtes GIF

Files

File	Size	Argmax agreement	Max prob gap	Sec/request	Peak RAM	Load setting
Q4_K_M	2.71 GB	97.5%	0.3419	10.03 s	3.03 GB	no mmap, repack on (jevalt default)
Q5_K_M	3.08 GB	95.0%	0.1494	17.33 s	3.39 GB	no mmap, repack on (jevalt default)
Q8_0	4.48 GB	97.5%	0.0315	11.58 s	4.76 GB	no mmap, repack on (jevalt default)

Measured by the export job on an HF cpu-upgrade machine (8 vCPU): argmax agreement and probability gap against the full-precision model on held-out validation requests, seconds per request with 8 threads, peak RAM of a fresh process serving the file at 4k context with 4 threads. The load setting column shows the llama.cpp configuration behind the RAM figure.

Use with jevalt

pip install "jevalt[serve,gguf] @ git+https://github.com/mertkayacs/jevalt"
jevalt serve --model mertkayacs/Wahler-4B-GGUF --file Wahler-4B-Q4_K_M.gguf

Then send the Jev request body to http://127.0.0.1:8000/v1/systemone. The answers come from the model's next-token probabilities at each decision marker, which the jevalt server reads through llama-cpp-python. llama.cpp's own llama-server and LM Studio can load the file, but their chat endpoints return generated text, so they give you neither the Jev API nor calibrated probabilities.