Open-source multilingual part-of-speech tagging model: xlm-roberta-base-ft-udpos28-gd - Supports tagging for Scottish Gaelic and other languages

Xlm Roberta Base Ft Udpos28 Gd

Developed by wietsedv

A multilingual POS tagging model based on XLM-RoBERTa, supporting POS tagging tasks for multiple languages including Scottish Gaelic.

Sequence Labeling

Transformers

OtherOpen Source License:Apache-2.0 #Multilingual POS tagging #Scottish Gaelic optimization #Cross-lingual transfer

Downloads 14

Release Time : 3/2/2022

Model Overview

This model is a multilingual POS tagging model based on the XLM-RoBERTa architecture, fine-tuned specifically for multiple languages in Universal Dependencies v2.8, with special optimization for Scottish Gaelic POS tagging performance.

Model Features

Multilingual support

Supports POS tagging tasks for over 100 languages.

High performance

Excellent performance across multiple languages, especially achieving 93.3% accuracy for Scottish Gaelic.

Based on XLM-RoBERTa

Fine-tuned using the powerful XLM-RoBERTa base model.

Model Capabilities

POS tagging

Multilingual text processing

Use Cases

Natural Language Processing

Text analysis

Perform POS tagging on texts in multiple languages.

Accurately identifies word POS categories.

Linguistic research

Supports research and analysis of rare languages.

Provides accurate POS tagging results.

🚀 XLM-RoBERTa base Universal Dependencies v2.8 POS tagging: Scottish Gaelic

This model addresses the task of part - of - speech tagging across multiple languages, leveraging the power of the XLM - RoBERTa base architecture. It provides valuable insights into cross - lingual transfer, as detailed in the associated research paper.

🚀 Quick Start

This model is part of our paper called:

Make the Best of Cross - lingual Transfer: Evidence from POS Tagging with over 100 Languages

Check the Space for more details.

💻 Usage Examples

Basic Usage

from transformers import AutoTokenizer, AutoModelForTokenClassification

tokenizer = AutoTokenizer.from_pretrained("wietsedv/xlm-roberta-base-ft-udpos28-gd")
model = AutoModelForTokenClassification.from_pretrained("wietsedv/xlm-roberta-base-ft-udpos28-gd")

📄 License

The model is released under the Apache - 2.0 license.

📚 Documentation

Model Information

Property	Details
Model Type	xlm - roberta - base - ft - udpos28 - gd
Tags	part - of - speech, token - classification
Datasets	universal_dependencies
Metrics	accuracy

Results

The model's performance on different languages in the Universal Dependencies v2.8 dataset is as follows:

Language	Test Accuracy
English	75.0
Dutch	77.8
German	76.5
Italian	70.8
French	74.6
Spanish	78.7
Russian	79.2
Swedish	78.9
Norwegian	72.7
Danish	78.0
Low Saxon	51.0
Akkadian	47.0
Armenian	69.2
Welsh	77.0
Old East Slavic	70.1
Albanian	76.1
Slovenian	64.3
Guajajara	42.6
Kurmanji	73.6
Turkish	71.7
Finnish	74.4
Indonesian	74.2
Ukrainian	78.7
Polish	81.4
Portuguese	77.9
Kazakh	73.3
Latin	68.8
Old French	48.7
Buryat	58.4
Kaapor	24.6
Korean	58.9
Estonian	76.8
Croatian	74.0
Gothic	29.4
Swiss German	48.3
Assyrian	20.1
North Sami	44.3
Naija	40.4
Latvian	76.7
Chinese	51.6
Tagalog	68.3
Bambara	30.3
Lithuanian	77.2
Galician	77.6
Vietnamese	56.5
Greek	79.1
Catalan	74.5
Czech	78.7
Erzya	51.6
Bhojpuri	49.4
Thai	57.1
Marathi	72.4
Basque	65.9
Slovak	80.3
Kiche	45.0
Yoruba	32.5
Warlpiri	43.7
Tamil	76.7
Maltese	34.9
Ancient Greek	59.3
Icelandic	73.1
Mbya Guarani	34.5
Urdu	56.0
Romanian	74.4
Persian	77.3
Apurina	48.4
Japanese	38.6
Hungarian	78.5
Hindi	60.5
Classical Chinese	31.6
Komi Permyak	50.4
Faroese	71.2
Sanskrit	33.5
Livvi	61.6
Arabic	81.6
Wolof	38.1
Bulgarian	76.6
Akuntsu	39.8
Makurap	23.3
Kangri	44.0
Breton	60.9
Telugu	74.5
Cantonese	48.9
Old Church Slavonic	47.7
Karelian	65.4
Upper Sorbian	70.9
South Levantine Arabic	68.4
Komi Zyrian	45.0
Irish	76.6
Nayini	44.9
Munduruku	34.0
Manx	52.0
Skolt Sami	39.7
Afrikaans	74.0
Old Turkish	37.1
Tupinamba	48.1
Belarusian	79.7
Serbian	72.7
Moksha	49.3
Western Armenian	68.1
Scottish Gaelic	93.3
Khunsari	44.6
Hebrew	86.5
Uyghur	67.5
Chukchi	38.8

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご