Logo Lanfrica

EngAmrAli/Egypt-Tourism-chatbot

Domaine:

natural language processing

Type de record:

modelsoftware
Créateur:
Eng
Hôte:
Egypt Tourism AI Assistant is a multimodal web application that answers questions about Egyptian landmarks and generates landmark images from user prompts. It uses a Qwen2.5 LoRA chatbot fine-tuned on a handmade Egypt tourism dataset and a Conditional GAN trained on an Egypt landmarks image dataset, with a simple Python API # Egypt Tourism AI Assistant A multimodal AI web application for Egypt tourism. The system can answer tourism questions about Egyptian landmarks and generate landmark images from prompts such as `generate pyramid image`. ## Project Overview This project combines two AI models in one simple web assistant: 1. **Tourism Chatbot** - Uses a Qwen2.5 language model fine-tuned with LoRA. - Answers questions about Egyptian tourism, landmarks, locations, and travel tips. - Trained on a handmade text dataset created for this project because a suitable public Egypt tourism chatbot dataset was not available. 2. **Landmark Image Generator** - Uses a Conditional GAN model. - Generates images for supported Egyptian landmark classes. - Trained on the Egypt Landmarks Kaggle dataset. The frontend is built with **HTML, CSS, and JavaScript**, and the backend API is built with **Python**. ## Features - Chatbot for Egypt tourism questions. - Image generation from prompts. - Automatic prompt routing: - Normal questions go to the chatbot. - Image requests go to the CGAN model. - Simple web interface. - Python API endpoint for frontend integration. - Dataset fallback answers if the Qwen model cannot be loaded locally. ## Example Prompts Text questions: ```text Tell me about the Pyramids of Giza. Where is Siwa located? Give me travel tips for Aswan. Why should I visit the Great Sphinx? ``` Image generation: ```text generate pyramid image generate Bent Pyramid image generate Siwa image create Temple of Kom Ombo picture ``` ## Technologies Used - Python - HTML - CSS - JavaScript - PyTorch - Transformers - PEFT / LoRA - Qwen2.5-0.5B-Instruct - Conditional GAN - Pillow ## Project Structure ```text AI_Application_Project/ │ ├── app.py ├── requirements.txt ├── README.md │ ├── web/ │ ├── index.html │ ├── styles.css │ └── script.js │ ├── data/ │ ├── egypt_tourism_gpt2_plain_text.txt │ ├── egypt-tourism-qwen-lora/ │ └── egypt-tourism-qwen-lora/ │ ├── adapter_config.jso …

Languages

Licenses