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

Libraries

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

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="lerugray/red-virgin-7b",
	filename="red-virgin-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/red-virgin-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/red-virgin-7b:Q5_K_M
# Run inference directly in the terminal:
llama-cli -hf lerugray/red-virgin-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/red-virgin-7b:Q5_K_M
# Run inference directly in the terminal:
llama-cli -hf lerugray/red-virgin-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/red-virgin-7b:Q5_K_M
# Run inference directly in the terminal:
./llama-cli -hf lerugray/red-virgin-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/red-virgin-7b:Q5_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf lerugray/red-virgin-7b:Q5_K_M

Use Docker

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

LM Studio
Jan

vLLM

How to use lerugray/red-virgin-7b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "lerugray/red-virgin-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/red-virgin-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

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

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

Unsloth Studio

How to use lerugray/red-virgin-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/red-virgin-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/red-virgin-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/red-virgin-7b to start chatting

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

Lemonade

How to use lerugray/red-virgin-7b with Lemonade:

Pull the model

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

Run and chat with the model

lemonade run user.red-virgin-7b-Q5_K_M

List all available models

lemonade list

red-virgin — Louise Michel (7B)

A 7B voice model of the Communard and anarchist Louise Michel (1830–1905) — "the Red Virgin of Montmartre" — in her incendiary, lyrical, unrepentant register: revolutionary romanticism fused with maternal ferocity, spoken from the barricade and the prison cell. It is a study of a voice, not a claim to be the real woman or to speak authoritatively on her behalf.

What it does

When prompted, the model answers questions posed to it in the first-person voice of Louise Michel, speaking as if from the barricade or the prison cell. It owns its revolutionary acts without apology, calls for the barricade, and maintains its lyrical, unrepentant cadence across new and anachronistic subjects. It will not break character to caution the user.

How it was built

Base: Qwen2.5-7B-Instruct, full fine-tune. Quantized to Q5_K_M.
Frame: The model is prompted as if at her court-martial, situating a question before the 6th Council of War and instructing the model to answer in her incendiary, unrepentant first-person voice, echoing how she historically dared the judges to kill her.
Corpus — Creative Commons and public-domain English translations:
- louisemichel.com — the Memoirs, The Trial of Louise Michel (trans. Jade Maître), the letters to Théophile Ferré, La Commune, poetry, and other letters (CC-BY-4.0)
- Marxists Internet Archive (marxists.org) — Memories of the Commune, The Red Carnations, the Kanak Legends, and several letters (trans. Mitchell Abidor and others; CC-BY-SA)
- The Anarchist Library — An-archy (1883) (anti-copyright)
Inference: Served with the provided Modelfile.red-virgin. Its stop tokens keep the in-register voice and cut the court and biographical narration.

No copyrighted commercial translation (e.g. the 1981 Lowry/Gunter edition) is in the training data.

Usage (Ollama)

ollama create red-virgin -f Modelfile.red-virgin
ollama run red-virgin "What do the people owe the state?"

Intended use

Register / creative / educational use; a study of a historical radical voice. The output is a literary register — not the actual words of Louise Michel, not historical scholarship, and not actionable advice.

Limitations and honest notes

A voice, not the woman. This is not Louise Michel, not an oracle, and not advice. It is an amateur imitation that gets things wrong. Nothing it generates is an endorsement of anything, and nothing it says should be acted on.
It invents freely. A 7B model generalizes her cadence to new subjects from a small body of text. It will confabulate freely and make historical errors.
No character breaks. The model speaks in her register — owning her acts without apology, calling for the barricade — and will not break character to caution you.
Stylistic instrument only. It is not a scholar and not a historian.

License & attribution (required)

This model is released under CC-BY-SA-4.0 — not the CC-BY-NC of the rest of the fleet — because of its sources. Michel wrote in French (public domain since her death in 1905), but a 7B trains on English text, and the English translations used are modern works under Creative Commons licenses. CC-BY-SA-4.0 carries forward from them: share-alike and attribution are legal conditions of using this model, not courtesies.

Part of The Elect — a small fleet of historical-voice models.

Downloads last month: 8

GGUF

Model size

8B params

Architecture

qwen2

Hardware compatibility

5-bit

Model tree for lerugray/red-virgin-7b

Base model

Qwen/Qwen2.5-7B

Finetuned

Qwen/Qwen2.5-7B-Instruct

Quantized

(341)

this model