Instructions to use uaritm/gemmamed_cardio with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use uaritm/gemmamed_cardio with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="uaritm/gemmamed_cardio")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("uaritm/gemmamed_cardio")
model = AutoModelForImageTextToText.from_pretrained("uaritm/gemmamed_cardio")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

llama-cpp-python

How to use uaritm/gemmamed_cardio with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="uaritm/gemmamed_cardio",
	filename="gemma_ua_med_final_q4km.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Inference
Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use uaritm/gemmamed_cardio with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf uaritm/gemmamed_cardio
# Run inference directly in the terminal:
llama-cli -hf uaritm/gemmamed_cardio

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf uaritm/gemmamed_cardio
# Run inference directly in the terminal:
llama-cli -hf uaritm/gemmamed_cardio

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 uaritm/gemmamed_cardio
# Run inference directly in the terminal:
./llama-cli -hf uaritm/gemmamed_cardio

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 uaritm/gemmamed_cardio
# Run inference directly in the terminal:
./build/bin/llama-cli -hf uaritm/gemmamed_cardio

Use Docker

docker model run hf.co/uaritm/gemmamed_cardio

LM Studio
Jan

vLLM

How to use uaritm/gemmamed_cardio with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/uaritm/gemmamed_cardio

SGLang

How to use uaritm/gemmamed_cardio with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "uaritm/gemmamed_cardio" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "uaritm/gemmamed_cardio",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "uaritm/gemmamed_cardio" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "uaritm/gemmamed_cardio",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use uaritm/gemmamed_cardio with Ollama:
```
ollama run hf.co/uaritm/gemmamed_cardio
```

Unsloth Studio new

How to use uaritm/gemmamed_cardio 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 uaritm/gemmamed_cardio 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 uaritm/gemmamed_cardio to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for uaritm/gemmamed_cardio to start chatting

Docker Model Runner
How to use uaritm/gemmamed_cardio with Docker Model Runner:
```
docker model run hf.co/uaritm/gemmamed_cardio
```

Lemonade

How to use uaritm/gemmamed_cardio with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull uaritm/gemmamed_cardio

Run and chat with the model

lemonade run user.gemmamed_cardio-{{QUANT_TAG}}

List all available models

lemonade list

🇺🇦 Gemma-UA-Cardio-Q4KM (Specialized Ukrainian LLM)

🩺 High-Quality Cardiology Assistant (Ukrainian Language)

This model is a highly specialized, instruction-following version of the Gemma-4B-It model (based on the size of 2.4GB, likely the 8B or a heavily compressed version), meticulously fine-tuned for providing cardiology-related information and answering medical queries in Ukrainian.

The adaptation process involved a crucial two-stage LoRA fine-tuning approach:

Linguistic Adaptation on a large general Ukrainian corpus (14500 cardiological epicrises).
Domain Specialization on a dedicated corpus of cardiovascular health and clinical data.

💾 Model Details & Files

This repository contains the highly optimized GGUF file, ready for immediate, efficient inference on consumer hardware.

Detail	Value
Base Model	google/gemma-4b-it
Language	Ukrainian (Specialized)
Specialization	Cardiology (Cardiovascular medicine, clinical terminology)
Quantization	GGUF Q4KM (Highly efficient)
GGUF File Size	~2.4 GB
Context Length	4096 (Recommended minimum)
Pipeline Tool	`llama.cpp` for GGUF conversion/quantization

Downloadable Files

gemma_ua_med_final_q8.gguf: The main file, ready for use with llama.cpp, Ollama, or LM Studio.

🇺🇦 Gemma-UA-Cardio-Q4KM (Specialized Ukrainian LLM)

🩺 High-Quality Cardiology Assistant (Ukrainian Language)

This model is a highly specialized, instruction-following version of the Gemma-4B-Instruct model (based on the size of 2.4GB, likely the 4B or a heavily compressed version), meticulously fine-tuned for providing cardiology-related information and answering medical queries in Ukrainian.

The adaptation process involved a crucial two-stage LoRA fine-tuning approach: 1. Linguistic Adaptation on a large general Ukrainian corpus. 2. Domain Specialization on a dedicated corpus of cardiovascular health and clinical data.

💾 Model Details & Files

This repository contains the highly optimized GGUF file, ready for immediate, efficient inference on consumer hardware.

Detail	Value
Base Model	google/gemma-4b-it
Language	Ukrainian (Specialized)
Specialization	Cardiology (Cardiovascular medicine, clinical terminology)
Quantization	GGUF Q4KM (Highly efficient)
GGUF File Size	~2.4 GB
Context Length	4096 (Recommended minimum)
Pipeline Tool	`llama.cpp` for GGUF conversion/quantization

Downloadable Files

gemma_ua_med_final_q4km.gguf: The main file, ready for use with llama.cpp, Ollama, or LM Studio.

🚀 How to Run (via llama.cpp CLI)

The GGUF format is ideal for running on CPU-dominant systems or systems with smaller VRAM using the llama.cpp framework.

1. Prerequisites

Ensure you have llama.cpp compiled. You will use the llama-cli binary.

2. Command Line Interface (CLI)

Use the following command to run the model in interactive chat mode. The System Prompt (-sys) is essential for activating the cardiology persona.

3. Example Usage (Ukrainian) This command assumes you are running from Ollama (Modelfile)

FROM "./gemma_ua_med_final_q4km.gguf"

SYSTEM """ Ви — лікар-кардіолог. На основі даних пацієнта та виписного епікризу сформуйте клінічно обґрунтовані рекомендації. """

TEMPLATE """user {{ .System }}

{{ .Prompt }} model """

--- IMPORTANT: Stop Tokens ---

PARAMETER stop "" PARAMETER stop "user" PARAMETER stop "model" PARAMETER stop "system"

--- Recommended generation parameters ---

PARAMETER temperature 0.6 PARAMETER top_p 0.9 PARAMETER num_ctx 4096 PARAMETER repeat_penalty 1.15 PARAMETER repeat_last_n 256

Key Parameters: -m: Specifies the path to the GGUF model file.

-sys: System Prompt—sets the model's professional role and required language.

-t $(nproc): Utilizes all available CPU cores for maximum speed.

-i: Activates interactive chat mode.

4. Limitations

Optimized for cardiology and cardiac surgery
Reduced accuracy outside these domains
No vision capabilities (text-only MedGemma IT)
May generate incomplete or generalized recommendations

5. Citing & Authors

If you use this model in your research, please cite:

@misc{Ostashko2025MedGemmaCardiology, title = {MedGemma-4B-Cardiology: A Domain-Finetuned Clinical LLM for Cardiology}, author = {Uaritm}, year = {2025}, url = {ai.esemi.org} }

Project homepage: https://ai.esemi.org

LicenseThe use of this model is subject to the terms of the original Gemma License. Please review and adhere to the associated licensing terms for the base model.

Downloads last month: 61

Safetensors

Model size

4B params

Tensor type

BF16

Model tree for uaritm/gemmamed_cardio

Base model

google/gemma-3-4b-pt

Finetuned

google/gemma-3-4b-it

Quantized

(218)

this model