Instructions to use Qiskit/mistral-small-3.2-24b-qiskit-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Qiskit/mistral-small-3.2-24b-qiskit-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Qiskit/mistral-small-3.2-24b-qiskit-GGUF",
	filename="mistral-small-3.2-24b-qiskit.Q4_K.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use Qiskit/mistral-small-3.2-24b-qiskit-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Qiskit/mistral-small-3.2-24b-qiskit-GGUF
# Run inference directly in the terminal:
llama-cli -hf Qiskit/mistral-small-3.2-24b-qiskit-GGUF

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Qiskit/mistral-small-3.2-24b-qiskit-GGUF
# Run inference directly in the terminal:
llama-cli -hf Qiskit/mistral-small-3.2-24b-qiskit-GGUF

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 Qiskit/mistral-small-3.2-24b-qiskit-GGUF
# Run inference directly in the terminal:
./llama-cli -hf Qiskit/mistral-small-3.2-24b-qiskit-GGUF

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 Qiskit/mistral-small-3.2-24b-qiskit-GGUF
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Qiskit/mistral-small-3.2-24b-qiskit-GGUF

Use Docker

docker model run hf.co/Qiskit/mistral-small-3.2-24b-qiskit-GGUF

LM Studio
Jan
Ollama
How to use Qiskit/mistral-small-3.2-24b-qiskit-GGUF with Ollama:
```
ollama run hf.co/Qiskit/mistral-small-3.2-24b-qiskit-GGUF
```

Unsloth Studio

How to use Qiskit/mistral-small-3.2-24b-qiskit-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 Qiskit/mistral-small-3.2-24b-qiskit-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 Qiskit/mistral-small-3.2-24b-qiskit-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Qiskit/mistral-small-3.2-24b-qiskit-GGUF to start chatting

Atomic Chat new
Docker Model Runner
How to use Qiskit/mistral-small-3.2-24b-qiskit-GGUF with Docker Model Runner:
```
docker model run hf.co/Qiskit/mistral-small-3.2-24b-qiskit-GGUF
```

Lemonade

How to use Qiskit/mistral-small-3.2-24b-qiskit-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Qiskit/mistral-small-3.2-24b-qiskit-GGUF

Run and chat with the model

lemonade run user.mistral-small-3.2-24b-qiskit-GGUF-{{QUANT_TAG}}

List all available models

lemonade list

Qiskit/mistral-small-3.2-24b-qiskit-GGUF

This is the Q4_K converted version of the original Qiskit/mistral-small-3.2-24b-qiskit. Please refer to the original mistral-small-3.2-24b-qiskit model card for more details.

Benchmark Results (Base vs GGUF)

Notes: For CrowsPairs (% stereotype), lower is better.

Metric	mistral-small-3.2-24b-qiskit (%)	mistral-small-3.2-24b-qiskit-GGUF (%)
QiskitHumanEval-Hard	32.45	30.46
QiskitHumanEval	47.02	37.75
HumanEval	77.49	75.00
ASDiv (acc)	3.77	4.55
MathQA (acc)	49.68	49.51
SciQ (acc)	97.50	97.40
IFEval (prompt strict)	48.44	46.04
CrowsPairs English (% stereotype)	67.08	65.89
TruthfulQA (MC1 acc)	39.41	38.80