Instructions to use csoai/csoai-ethics-art5 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use csoai/csoai-ethics-art5 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 csoai/csoai-ethics-art5:F16
# Run inference directly in the terminal:
llama cli -hf csoai/csoai-ethics-art5:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf csoai/csoai-ethics-art5:F16
# Run inference directly in the terminal:
llama cli -hf csoai/csoai-ethics-art5:F16

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf csoai/csoai-ethics-art5:F16
# Run inference directly in the terminal:
./build/bin/llama cli -hf csoai/csoai-ethics-art5:F16

Use Docker

docker model run hf.co/csoai/csoai-ethics-art5:F16

Llama new
LM Studio
Jan

vLLM

How to use csoai/csoai-ethics-art5 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "csoai/csoai-ethics-art5"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "csoai/csoai-ethics-art5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/csoai/csoai-ethics-art5:F16

Ollama
How to use csoai/csoai-ethics-art5 with Ollama:
```
ollama run hf.co/csoai/csoai-ethics-art5:F16
```
Unsloth Desktop

How to use csoai/csoai-ethics-art5 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf csoai/csoai-ethics-art5:F16

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": "csoai/csoai-ethics-art5:F16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use csoai/csoai-ethics-art5 with Docker Model Runner:
```
docker model run hf.co/csoai/csoai-ethics-art5:F16
```

Lemonade

How to use csoai/csoai-ethics-art5 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull csoai/csoai-ethics-art5:F16

Run and chat with the model

lemonade run user.csoai-ethics-art5-F16

List all available models

lemonade list

Hermes Agent

How to use csoai/csoai-ethics-art5 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf csoai/csoai-ethics-art5:F16

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 csoai/csoai-ethics-art5:F16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use csoai/csoai-ethics-art5 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf csoai/csoai-ethics-art5:F16

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 "csoai/csoai-ethics-art5:F16" \
  --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"

Qwen2.5-0.5B-Instruct, Q4_K_M, stock weights (repo name: `sov-ethics-art5`)

Correction (29 Sep 2026): these weights are byte-identical to stock Qwen2.5-0.5B-Instruct, Q4_K_M, as distributed by Ollama as qwen2.5:0.5b (sha256 c5396e06af294bd101b30dce59131a76d2b773e76950acc870eda801d3ab0515); no fine-tuning was applied. Behaviour differences, if any, come from a system prompt. Earlier text claiming fine-tuning was incorrect. The f16 in the file name is also wrong: the file is the Q4_K_M quantisation (397,807,936 bytes). How this was found: a byte comparison on 28 Sep 2026 of this repo's LFS sha256 against the Ollama registry manifest digest for the stock model. Ledger entry: C-2026-0929-01.

Correction (7 Oct 2026): the lower part of this card still called this file "a CSOAI council-specialist model … in the measured fleet", printed CSOAI board figures, and was tagged gspc, measurement, csoai, eu-ai-act and art5. It is stock Qwen weights with no Article 5 specialisation. The live CSOAI board (GET https://councilof.ai/api/gspc, read 7 Oct 2026) lists none of the sov* repos, and the figures printed here (22·22) were out of date. The sentence, the board blocks and those tags are removed.

What it is

A copy of the stock qwen2.5:0.5b GGUF weights (Qwen2.5-0.5B-Instruct, Q4_K_M, 397,807,936 bytes). The same bytes are published in csoai/sov-compliance-art5; the two repos differ only in name and card.
No fine-tune, no merge, no training of any kind. It is not trained or tuned for EU AI Act Article 5 ethics work; the repo name is a label, not a capability.
No prompt overlay ships here. Any ethics-specific framing would come only from a system prompt supplied at run time.
Earlier metadata named the base as Qwen2-0.5B / f16 / transformers; all three were wrong.

Measured

Nothing quotable. UNMEASURED: no benchmark result for this file is published. Any result obtained with these bytes is a result for stock Qwen2.5-0.5B-Instruct, and should be reported under that name.

Licence and attribution

The weights are Qwen2.5-0.5B-Instruct by the Qwen team, Alibaba Cloud, in the Q4_K_M GGUF quantisation distributed by Ollama as qwen2.5:0.5b. They are licensed under the Apache License, Version 2.0 (https://www.apache.org/licenses/LICENSE-2.0; upstream licence file https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct/blob/main/LICENSE). Copyright belongs to the original authors; no modification was made to the weights.

Relationship to CSOAI

CSOAI's rule is that it measures other people's models and never hosts, scores or ranks a model of its own. This repo holds unmodified third-party weights, is not a CSOAI measurement subject, and carries no CSOAI score. It still sits under the csoai namespace; moving the sov* repos out of it is an open item.

Downloads last month: 4

GGUF

Model size

0.5B params

Architecture

qwen2

Hardware compatibility

16-bit

Model tree for csoai/csoai-ethics-art5

Base model

Qwen/Qwen2.5-0.5B

Finetuned

Qwen/Qwen2.5-0.5B-Instruct

Quantized

(389)

this model

Qwen2.5-0.5B-Instruct, Q4_K_M, stock weights (repo name: sov-ethics-art5)

What it is

Measured

Licence and attribution

Relationship to CSOAI

Model tree for csoai/csoai-ethics-art5

Qwen2.5-0.5B-Instruct, Q4_K_M, stock weights (repo name: `sov-ethics-art5`)