Instructions to use lerugray/clausewitz-7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use lerugray/clausewitz-7b with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="lerugray/clausewitz-7b",
	filename="clausewitz-qwen2-5-7b-instruct-Q5_K_M.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use lerugray/clausewitz-7b with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf lerugray/clausewitz-7b:Q5_K_M
# Run inference directly in the terminal:
llama-cli -hf lerugray/clausewitz-7b:Q5_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf lerugray/clausewitz-7b:Q5_K_M
# Run inference directly in the terminal:
llama-cli -hf lerugray/clausewitz-7b:Q5_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 lerugray/clausewitz-7b:Q5_K_M
# Run inference directly in the terminal:
./llama-cli -hf lerugray/clausewitz-7b:Q5_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 lerugray/clausewitz-7b:Q5_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf lerugray/clausewitz-7b:Q5_K_M

Use Docker

docker model run hf.co/lerugray/clausewitz-7b:Q5_K_M

LM Studio
Jan

vLLM

How to use lerugray/clausewitz-7b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "lerugray/clausewitz-7b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "lerugray/clausewitz-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/lerugray/clausewitz-7b:Q5_K_M

Ollama
How to use lerugray/clausewitz-7b with Ollama:
```
ollama run hf.co/lerugray/clausewitz-7b:Q5_K_M
```

Unsloth Studio

How to use lerugray/clausewitz-7b 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 lerugray/clausewitz-7b 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 lerugray/clausewitz-7b to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for lerugray/clausewitz-7b to start chatting

Atomic Chat new
Docker Model Runner
How to use lerugray/clausewitz-7b with Docker Model Runner:
```
docker model run hf.co/lerugray/clausewitz-7b:Q5_K_M
```

Lemonade

How to use lerugray/clausewitz-7b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull lerugray/clausewitz-7b:Q5_K_M

Run and chat with the model

lemonade run user.clausewitz-7b-Q5_K_M

List all available models

lemonade list

clausewitz: a dual-mode On War strategic model

A 7B tune of Carl von Clausewitz's On War (Vom Kriege) that does two things: it speaks in his analytical theoretical register, and it reasons — applying his apparatus to a situation you give it. Two serve frames ship with it:

discuss — the voice. Ask it about war, friction, the trinity, attack and defense, and it answers as the theorist.
advisor — the framework-reasoner. Hand it a situation — a conflict, a strategy, a market, a game position — and it analyzes it through On War's apparatus: the political object, the center of gravity (Schwerpunkt), friction and uncertainty, the paradoxical trinity, attack vs. defense, the culminating point of victory, moral forces, the genius of the commander.

What it does

discuss (voice):

What is war? "War has two principal objects, the destruction of the enemy's military force and his submission. Destruction means the physical elimination of an armed force; subjection means its loss of freedom of action…"

How would your theory apply to a war fought with machines and networks, where no soldier meets another? "The essence of war is nothing but the hostile action of human beings upon one another; the means therefore are always men. If we cannot see any individuals here, we must look for them behind the scenes."

advisor (framework-reasoner) — on a situation it was never trained on:

A small open-source AI project trying to displace a dominant commercial AI company. "The paradoxical trinity applies: passion drives the developers' commitment; chance governs which breakthroughs might tip the balance; reason must guide strategic decisions… Attack vs. defense: your project is inherently defensive — protecting an open platform against commercial encroachment… The culminating point of victory occurs when your network effect becomes strong enough… Moral forces are decisive… Judgment: your political object is clear, your center of gravity secure if you nurture trust."

How it was built

Base: Qwen2.5-7B-Instruct, full fine-tune.
Format: completion (raw text).
Corpus (1,199 records): On War in the public-domain Graham/Maude translation (Project Gutenberg #1946), chunked into 1,136 completion records (~293k words); plus 63 synthetic Mode-B worked-analyses — situations analyzed through the On War apparatus, generated and curated to teach the model to apply the framework rather than only recite it. The synthetic bridge is the difference between a voice and a reasoner.
Inference: two lead-in frames (discuss / advisor) elicit the two modes.

Usage (Ollama)

ollama create clausewitz         -f Modelfile.clausewitz-discuss
ollama create clausewitz-advisor -f Modelfile.clausewitz-advisor

ollama run clausewitz "Which is the stronger form of war, attack or defense?"
ollama run clausewitz-advisor "A tenant resisting eviction by a corporate landlord with expensive lawyers."

Intended use

Strategic-reasoning aid, wargame and scenario analysis, creative and educational use, a theory-lens sparring partner. The output is an analytical register and a structured way of thinking about a contest — not professional, legal, or military advice.

Limitations and honest notes

A lens, not an oracle. It reasons inside one framework; it will confidently map On War's concepts onto anything, including situations where they fit poorly. Treat it as a perspective.
Period diction, and it can confabulate; the On War text it draws on is the 19th-century Graham translation.
All-public-domain source (On War) + a small curated synthetic bridge; weights released for non-commercial use.

License

CC-BY-NC-4.0. On War is public domain; the weights are released for non-commercial use. No warranty.

Part of the Elect — a roster of public-domain voice and register models. Clausewitz is the strategic-reasoning member.

Downloads last month: 9

GGUF

Model size

8B params

Architecture

qwen2

Hardware compatibility

5-bit

Model tree for lerugray/clausewitz-7b

Base model

Qwen/Qwen2.5-7B

Finetuned

Qwen/Qwen2.5-7B-Instruct

Quantized

(341)

this model