T

T5 V1.1 Base Dutch Cnn Test

Developed by yhavinga
A Dutch news summarization model based on the T5 architecture, fine-tuned for the Dutch version of CNN Daily Mail
Downloads 176
Release Time : 3/2/2022

Model Overview

This model is a sequence-to-sequence model fine-tuned on the Dutch T5 base model, specifically designed for generating summaries of Dutch news articles.

Model Features

Dutch Language Specialization
Trained on the cleaned Dutch mC4 dataset, specifically optimized for Dutch text processing
High-Quality Summaries
Fine-tuned on the Dutch CNN Daily Mail dataset, achieving ROUGE-L 25.9 summarization quality
Optimized Tokenizer
Uses a SentencePiece tokenizer specifically trained for Dutch, delivering better processing results
Data Cleaning
Training data underwent strict filtering to remove low-quality content and anomalies

Model Capabilities

Dutch Text Understanding
News Summarization
Long Text Compression

Use Cases

News Media
Automatic News Summarization
Automatically generates concise summaries for Dutch news articles
Average summary length of around 91 words, with ROUGE-L score of 25.9
Content Analysis
Key Information Extraction
Extracts core information from lengthy Dutch documents
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase