Instructions to use Mariem-Daha/RACHIDA-Mini-2.03B-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 Mariem-Daha/RACHIDA-Mini-2.03B-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 Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Mariem-Daha/RACHIDA-Mini-2.03B-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 Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Mariem-Daha/RACHIDA-Mini-2.03B-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 Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Mariem-Daha/RACHIDA-Mini-2.03B-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 Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use Mariem-Daha/RACHIDA-Mini-2.03B-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Mariem-Daha/RACHIDA-Mini-2.03B-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Mariem-Daha/RACHIDA-Mini-2.03B-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M

Ollama
How to use Mariem-Daha/RACHIDA-Mini-2.03B-GGUF with Ollama:
```
ollama run hf.co/Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M
```

Unsloth Studio

How to use Mariem-Daha/RACHIDA-Mini-2.03B-GGUF 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 Mariem-Daha/RACHIDA-Mini-2.03B-GGUF 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 Mariem-Daha/RACHIDA-Mini-2.03B-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Mariem-Daha/RACHIDA-Mini-2.03B-GGUF to start chatting

How to use Mariem-Daha/RACHIDA-Mini-2.03B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Mariem-Daha/RACHIDA-Mini-2.03B-GGUF with Docker Model Runner:
```
docker model run hf.co/Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M
```

Lemonade

How to use Mariem-Daha/RACHIDA-Mini-2.03B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.RACHIDA-Mini-2.03B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use Mariem-Daha/RACHIDA-Mini-2.03B-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 Mariem-Daha/RACHIDA-Mini-2.03B-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 Mariem-Daha/RACHIDA-Mini-2.03B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Mariem-Daha/RACHIDA-Mini-2.03B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Mariem-Daha/RACHIDA-Mini-2.03B-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 "Mariem-Daha/RACHIDA-Mini-2.03B-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"

RACHIDA-Mini-2.03B-Q4_K_M GGUF

This repository contains the GGUF selected for RACHIDA.AI, an offline maternal-health information and safety prototype submitted to the Africa Deep Tech Challenge 2026.

The file is a straight Q4_K_M quantization of Qwen/Qwen3-1.7B. No RACHIDA-specific fine-tune or adapter is merged into this artifact. RACHIDA.AI is the surrounding application and safety system; the model provenance remains Qwen3-1.7B.

File

File	Format	Size	SHA-256
`RACHIDA-Mini-2.03B-Q4_K_M.gguf`	GGUF Q4_K_M	1,282,439,040 bytes	`fc0d5cc4ce1cee6ec3566da40c48f9ed953e3ee2eedcce36e44b3c4924001ab9`

The GGUF tensor metadata reports 2,031,739,904 parameters. The upstream model is named Qwen3-1.7B; the 2.03B label records the tensor count reported by the converted artifact.

Intended use

Offline language-model evaluation with llama.cpp.
Preliminary maternal-health information and patient-education research.
Demonstration of CPU-only inference on commodity laptops.
Use inside an application where independently verified deterministic controls own safety-critical escalation decisions.

Important limitations

This artifact is not clinically validated. It is not a medical device, diagnostic system, triage system, prescription tool, or replacement for a clinician. Do not use it for autonomous diagnosis, treatment, prescribing, emergency-resource lookup, or decisions involving patient safety. The GGUF alone does not contain RACHIDA.AI's signed clinical content pack, deterministic danger-sign routing, output guard, or audit controls.

English is the only declared and evaluated language scope for this release. Do not provide patient records or other sensitive personal data to the model.

Run locally

llama-cli \
  -m RACHIDA-Mini-2.03B-Q4_K_M.gguf \
  -t 4 \
  -ngl 0 \
  -c 2048

The ADTC submission uses llama.cpp, four CPU threads, zero GPU layers, and no network access during inference.

Verified development measurements

Participant-mode measurements on the development laptop:

23.09 generated tokens/second;
2,521.14 ms first-token latency;
1,937.40 MB peak RSS;
0.76 ARC-Easy accuracy over 50 samples with seed 42.

These are development results, not official ADTC audit-machine results. The development host ran Windows and had 31.7 GB RAM; Ubuntu 22.04 / 8 GB results may differ. Core temperature was unavailable, so no thermal claim is made.

License and attribution

The weights retain the upstream Apache License 2.0. Base model: Qwen3-1.7B by the Qwen Team. Conversion, quantization, and CPU inference use llama.cpp, licensed under MIT.

Downloads last month: 95

GGUF

Model size

2B params

Architecture

qwen3

Hardware compatibility

4-bit

Model tree for Mariem-Daha/RACHIDA-Mini-2.03B-GGUF

Base model

Qwen/Qwen3-1.7B-Base

Finetuned

Qwen/Qwen3-1.7B

Quantized

(339)

this model