XLM-RoBERTa base Universal Dependencies v2.8 POS tagging: Irish

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

from transformers import AutoTokenizer, AutoModelForTokenClassification

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

Downloads last month: 4

Dataset used to train wietsedv/xlm-roberta-base-ft-udpos28-ga

Space using wietsedv/xlm-roberta-base-ft-udpos28-ga 1

Evaluation results

English Test accuracy on Universal Dependencies v2.8
self-reported

75.300
Dutch Test accuracy on Universal Dependencies v2.8
self-reported

79.500
German Test accuracy on Universal Dependencies v2.8
self-reported

76.200
Italian Test accuracy on Universal Dependencies v2.8
self-reported

73.600
French Test accuracy on Universal Dependencies v2.8
self-reported

76.400
Spanish Test accuracy on Universal Dependencies v2.8
self-reported

82.400
Russian Test accuracy on Universal Dependencies v2.8
self-reported

80.700
Swedish Test accuracy on Universal Dependencies v2.8
self-reported

78.700
Norwegian Test accuracy on Universal Dependencies v2.8
self-reported

74.500
Danish Test accuracy on Universal Dependencies v2.8
self-reported

77.900
Low Saxon Test accuracy on Universal Dependencies v2.8
self-reported

49.000
Akkadian Test accuracy on Universal Dependencies v2.8
self-reported

46.200
Armenian Test accuracy on Universal Dependencies v2.8
self-reported

71.800
Welsh Test accuracy on Universal Dependencies v2.8
self-reported

77.900
Old East Slavic Test accuracy on Universal Dependencies v2.8
self-reported

67.300
Albanian Test accuracy on Universal Dependencies v2.8
self-reported

79.800