Turkish Llava Med Pipeline V1.5 Mistral 7b

🚀 Turkish LLaVA-Med Pipeline (v1.5 - Mistral 7B)

This model is based on the LLaVA-Med v1.5 (Mistral 7B) architecture developed by Microsoft. It is customized to perform visual question-answering (VQA) on medical images with Turkish language support.

✨ Features

• Image-based analysis (X-ray, MR, etc.)
• Support for Turkish questions
• Generation of Turkish answers (with a translation layer)
• Support for a demo interface on Hugging Face (Gradio)

🚀 Quick Start

Starter Colab Notebook: https://colab.research.google.com/drive/1HsWo9g18r1EfKFMh0RcgZgrotoIIcThK?usp=sharing

from transformers import AutoModelForCausalLM, AutoTokenizer, CLIPImageProcessor
import torch

model = AutoModelForCausalLM.from_pretrained("nezahatkorkmaz/turkish-llava-med-pipeline-v1.5-mistral-7b", torch_dtype=torch.float16).to("cuda")
tokenizer = AutoTokenizer.from_pretrained("nezahatkorkmaz/turkish-llava-med-pipeline-v1.5-mistral-7b")
image_processor = CLIPImageProcessor.from_pretrained("nezahatkorkmaz/turkish-llava-med-pipeline-v1.5-mistral-7b")

💻 Usage Examples

Basic Usage

from transformers import AutoModelForCausalLM, AutoTokenizer, CLIPImageProcessor
import torch

model = AutoModelForCausalLM.from_pretrained("nezahatkorkmaz/turkish-llava-med-pipeline-v1.5-mistral-7b", torch_dtype=torch.float16).to("cuda")
tokenizer = AutoTokenizer.from_pretrained("nezahatkorkmaz/turkish-llava-med-pipeline-v1.5-mistral-7b")
image_processor = CLIPImageProcessor.from_pretrained("nezahatkorkmaz/turkish-llava-med-pipeline-v1.5-mistral-7b")

📋 Example Question

Image: Chest X-ray Question: "Is there any cardiomegaly in this image?" Answer: "Yes, there are signs of cardiomegaly in this image."

Developer: @nezahatkorkmaz
Original Model: microsoft/llava-med-v1.5-mistral-7b

📄 License

This project is licensed under the Apache-2.0 license.