AIbase
EN
Home
Punctuate All
P

Punctuate All

Developed by kredor
A multilingual punctuation prediction model fine-tuned based on xlm-roberta-base, supporting automatic punctuation completion for 12 European languages
Sequence Labeling
Transformers
Open Source License:MIT#Multilingual punctuation prediction#High-precision punctuation restoration#European Parliament corpus fine-tuning
Downloads 728.70k
Release Time : 4/9/2022

Model Overview

This model is used to automatically predict and complete punctuation in text, particularly suitable for punctuation restoration scenarios after speech-to-text conversion. Compared to the original model, this version supports more languages while using a smaller base model.

Model Features

Multilingual Support
Supports punctuation prediction for 12 European languages, adding 8 more languages compared to the original model
Efficient Model
Uses xlm-roberta-base instead of the large version, reducing computational resource requirements while maintaining good performance
High Accuracy
Achieves F1 scores of 0.85-0.95 for common punctuation marks (e.g., periods, commas)

Model Capabilities

Automatic punctuation completion
Multilingual text processing
Post-processing for speech-to-text

Use Cases

Speech Transcription Enhancement
Automatic Meeting Minutes Punctuation
Automatically adds punctuation to speech recognition output without punctuation
Can accurately restore 95% of periods and 86% of commas
Text Preprocessing
Machine Translation Preprocessing
Adds punctuation to raw text without punctuation to improve translation quality
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase