Instructions to use nusoidal/Solenoid_1.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use nusoidal/Solenoid_1.1 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 nusoidal/Solenoid_1.1:Q4_K_M # Run inference directly in the terminal: llama cli -hf nusoidal/Solenoid_1.1:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf nusoidal/Solenoid_1.1:Q4_K_M # Run inference directly in the terminal: llama cli -hf nusoidal/Solenoid_1.1: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 nusoidal/Solenoid_1.1:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf nusoidal/Solenoid_1.1: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 nusoidal/Solenoid_1.1:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf nusoidal/Solenoid_1.1:Q4_K_M
Use Docker
docker model run hf.co/nusoidal/Solenoid_1.1:Q4_K_M
- LM Studio
- Jan
- Ollama
How to use nusoidal/Solenoid_1.1 with Ollama:
ollama run hf.co/nusoidal/Solenoid_1.1:Q4_K_M
- Unsloth Desktop
- Pi
How to use nusoidal/Solenoid_1.1 with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf nusoidal/Solenoid_1.1: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": "nusoidal/Solenoid_1.1:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use nusoidal/Solenoid_1.1 with Docker Model Runner:
docker model run hf.co/nusoidal/Solenoid_1.1:Q4_K_M
- Lemonade
How to use nusoidal/Solenoid_1.1 with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull nusoidal/Solenoid_1.1:Q4_K_M
Run and chat with the model
lemonade run user.Solenoid_1.1-Q4_K_M
List all available models
lemonade list
- Hermes Agent
How to use nusoidal/Solenoid_1.1 with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf nusoidal/Solenoid_1.1: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 nusoidal/Solenoid_1.1:Q4_K_M
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use nusoidal/Solenoid_1.1 with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf nusoidal/Solenoid_1.1: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 "nusoidal/Solenoid_1.1: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"
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Solenoid 1.1
Fine-tuned legal LLM for OHADA jurisprudence and Niger fiscal law
Solenoid is a family of large language models fine-tuned by Nusoidal from Qwen2.5 architectures on Niger's legal corpora, in French (majority). It has the ability to provide answers related to OHADA business law, tax code, and administrative procedures.
Model Details
- Base model: Qwen2.5-1.5B
- Fine-tuning: LoRA (r=16, alpha=32) with ChatML prompt format
- Quantization: Q4_K_M (941 MB GGUF)
- Context window: 32,768 tokens
- Languages: French (primary), legal terminology
- License: Apache 2.0
Capabilities
Solenoid is specialized in:
- OHADA uniform acts (commercial law, company law, securities)
- Niger tax code and fiscal procedures
- Business registration and regulatory compliance
- Public procurement and administrative law
- Labor law and social security (CNSS)
Benchmarks
428 questions across OHADA, tax, labor, procurement, banking, civil, digital, environment, and real-world scenarios.
| Model | OHADA Refs | Niger Refs | Citations | Domain Score |
|---|---|---|---|---|
| Qwen2.5-1.5B (base) | 0.2 | 0.8 | 0.2 | 1.3 |
| Solenoid 1.1 | 0.7 | 1.1 | 1.7 | 3.5 |
Solenoid 1.1 outperforms Qwen2.5-1.5B (base) by 175% on domain relevance.
Solenoid scores 2.8x higher than the base model on domain-specific legal references. On OHADA questions specifically, the gap is even wider, with Solenoid consistently citing the correct uniform acts while the base model defaults to generic French law.
Benchmarked on RTX 5090 (NVIDIA Neptune), temperature 0.0, 256 max_tokens. Full code in benchmark_solenoid.py.
Prompt Format
Solenoid uses the ChatML format:
<|im_start|>system
Vous etes Solenoid, un assistant juridique specialise en droit OHADA.
<|im_end|>
<|im_start|>user
Quelles sont les etapes pour creer une SARL au Niger ?
<|im_end|>
<|im_start|>assistant
Files
| File | Description |
|---|---|
Solenoid_1.1-Q4_K_M.gguf |
Model weights (GGUF Q4_K_M, 941 MB) |
benchmark_solenoid.py |
Benchmark script (600 questions, GPU) |
benchmark_dashboard.py |
Interactive HTML dashboard generator |
benchmark_results/ |
Raw results + generated report + dashboard |
LICENSE |
Apache 2.0 |
Access
Solenoid 1.1 Q4_K_M is open source under Apache 2.0. Download the GGUF file and use with any llama.cpp-compatible inference engine.
# Download
wget https://huggingface.co/Nusoidal/Solenoid_1.1/resolve/main/Solenoid_1.1-Q4_K_M.gguf
# Run with llama.cpp
./llama-cli -m Solenoid_1.1-Q4_K_M.gguf -p "Qu'est-ce que l'OHADA ?" -n 256
For commercial licensing or other variants (Q8_0, Solenoid 1.2): nusoidal@gmail.com
License
Apache License 2.0. See LICENSE.
Solenoid is a trademark of Nusoidal.
- Downloads last month
- -
4-bit