Mt5 XLSUM Ua News
M
Mt5 XLSUM Ua News
Developed by yelyah
A headline generation model fine-tuned on Ukrainian news datasets based on the multilingual mT5 model, capable of generating concise and accurate headlines for Ukrainian news articles.
Downloads 110
Release Time : 5/20/2024
Model Overview
This model is specifically optimized for Ukrainian news content, implementing text summarization functionality through a sequence-to-sequence architecture, suitable for automated headline generation in the news media field.
Model Features
Ukrainian language specialization
Fine-tuned on the FIdo-AI/ua-news Ukrainian news dataset, with enhanced comprehension of Ukrainian news content
Multilingual foundation
Based on the mT5 multilingual pre-trained model, capable of cross-lingual knowledge transfer
Lightweight deployment
Supports quick deployment via the transformers library, with a batch size of 4 suitable for medium-configuration devices
Model Capabilities
Ukrainian text comprehension
News headline generation
Text summarization
Use Cases
News media
Automated news headline generation
Automatically generates headlines that capture key points of content for Ukrainian news platforms
ROUGE-1 score of 0.2452, capable of capturing key article information
News content summarization
Generates concise summaries of news articles
ROUGE-L score of 0.2348, maintaining semantic coherence with the original text
Featured Recommended AI Models