wietsedv's picture
update README
566c9f3
metadata
language:
  - ro
license: apache-2.0
library_name: transformers
tags:
  - part-of-speech
  - token-classification
datasets:
  - universal_dependencies
metrics:
  - accuracy
model-index:
  - name: xlm-roberta-base-ft-udpos28-ro
    results:
      - task:
          type: token-classification
          name: Part-of-Speech Tagging
        dataset:
          type: universal_dependencies
          name: Universal Dependencies v2.8
        metrics:
          - type: accuracy
            name: English Test accuracy
            value: 88.4
          - type: accuracy
            name: Dutch Test accuracy
            value: 86.1
          - type: accuracy
            name: German Test accuracy
            value: 87.3
          - type: accuracy
            name: Italian Test accuracy
            value: 88.2
          - type: accuracy
            name: French Test accuracy
            value: 91.3
          - type: accuracy
            name: Spanish Test accuracy
            value: 91.1
          - type: accuracy
            name: Russian Test accuracy
            value: 90.4
          - type: accuracy
            name: Swedish Test accuracy
            value: 90.7
          - type: accuracy
            name: Norwegian Test accuracy
            value: 85
          - type: accuracy
            name: Danish Test accuracy
            value: 91
          - type: accuracy
            name: Low Saxon Test accuracy
            value: 56.2
          - type: accuracy
            name: Akkadian Test accuracy
            value: 41.8
          - type: accuracy
            name: Armenian Test accuracy
            value: 88.4
          - type: accuracy
            name: Welsh Test accuracy
            value: 71.7
          - type: accuracy
            name: Old East Slavic Test accuracy
            value: 78.7
          - type: accuracy
            name: Albanian Test accuracy
            value: 90.2
          - type: accuracy
            name: Slovenian Test accuracy
            value: 80.3
          - type: accuracy
            name: Guajajara Test accuracy
            value: 39.3
          - type: accuracy
            name: Kurmanji Test accuracy
            value: 79.5
          - type: accuracy
            name: Turkish Test accuracy
            value: 79.5
          - type: accuracy
            name: Finnish Test accuracy
            value: 86
          - type: accuracy
            name: Indonesian Test accuracy
            value: 84.2
          - type: accuracy
            name: Ukrainian Test accuracy
            value: 89.7
          - type: accuracy
            name: Polish Test accuracy
            value: 89.5
          - type: accuracy
            name: Portuguese Test accuracy
            value: 90.3
          - type: accuracy
            name: Kazakh Test accuracy
            value: 85
          - type: accuracy
            name: Latin Test accuracy
            value: 81.8
          - type: accuracy
            name: Old French Test accuracy
            value: 65.7
          - type: accuracy
            name: Buryat Test accuracy
            value: 64.9
          - type: accuracy
            name: Kaapor Test accuracy
            value: 27.1
          - type: accuracy
            name: Korean Test accuracy
            value: 64.3
          - type: accuracy
            name: Estonian Test accuracy
            value: 87.5
          - type: accuracy
            name: Croatian Test accuracy
            value: 89.7
          - type: accuracy
            name: Gothic Test accuracy
            value: 35.1
          - type: accuracy
            name: Swiss German Test accuracy
            value: 55.5
          - type: accuracy
            name: Assyrian Test accuracy
            value: 16.8
          - type: accuracy
            name: North Sami Test accuracy
            value: 45
          - type: accuracy
            name: Naija Test accuracy
            value: 43.8
          - type: accuracy
            name: Latvian Test accuracy
            value: 89.5
          - type: accuracy
            name: Chinese Test accuracy
            value: 54.9
          - type: accuracy
            name: Tagalog Test accuracy
            value: 74
          - type: accuracy
            name: Bambara Test accuracy
            value: 32.9
          - type: accuracy
            name: Lithuanian Test accuracy
            value: 87.7
          - type: accuracy
            name: Galician Test accuracy
            value: 89.9
          - type: accuracy
            name: Vietnamese Test accuracy
            value: 66.2
          - type: accuracy
            name: Greek Test accuracy
            value: 88.9
          - type: accuracy
            name: Catalan Test accuracy
            value: 90
          - type: accuracy
            name: Czech Test accuracy
            value: 89.8
          - type: accuracy
            name: Erzya Test accuracy
            value: 51.5
          - type: accuracy
            name: Bhojpuri Test accuracy
            value: 55
          - type: accuracy
            name: Thai Test accuracy
            value: 64.9
          - type: accuracy
            name: Marathi Test accuracy
            value: 87.1
          - type: accuracy
            name: Basque Test accuracy
            value: 80.7
          - type: accuracy
            name: Slovak Test accuracy
            value: 89.8
          - type: accuracy
            name: Kiche Test accuracy
            value: 42.4
          - type: accuracy
            name: Yoruba Test accuracy
            value: 30.3
          - type: accuracy
            name: Warlpiri Test accuracy
            value: 46.2
          - type: accuracy
            name: Tamil Test accuracy
            value: 82.5
          - type: accuracy
            name: Maltese Test accuracy
            value: 38.3
          - type: accuracy
            name: Ancient Greek Test accuracy
            value: 67.8
          - type: accuracy
            name: Icelandic Test accuracy
            value: 85.1
          - type: accuracy
            name: Mbya Guarani Test accuracy
            value: 34.4
          - type: accuracy
            name: Urdu Test accuracy
            value: 63.4
          - type: accuracy
            name: Romanian Test accuracy
            value: 96.8
          - type: accuracy
            name: Persian Test accuracy
            value: 79
          - type: accuracy
            name: Apurina Test accuracy
            value: 43.1
          - type: accuracy
            name: Japanese Test accuracy
            value: 43.7
          - type: accuracy
            name: Hungarian Test accuracy
            value: 79.9
          - type: accuracy
            name: Hindi Test accuracy
            value: 70.6
          - type: accuracy
            name: Classical Chinese Test accuracy
            value: 40.8
          - type: accuracy
            name: Komi Permyak Test accuracy
            value: 57.2
          - type: accuracy
            name: Faroese Test accuracy
            value: 80.9
          - type: accuracy
            name: Sanskrit Test accuracy
            value: 40.4
          - type: accuracy
            name: Livvi Test accuracy
            value: 66.9
          - type: accuracy
            name: Arabic Test accuracy
            value: 83.5
          - type: accuracy
            name: Wolof Test accuracy
            value: 43.1
          - type: accuracy
            name: Bulgarian Test accuracy
            value: 91.2
          - type: accuracy
            name: Akuntsu Test accuracy
            value: 40.6
          - type: accuracy
            name: Makurap Test accuracy
            value: 20.5
          - type: accuracy
            name: Kangri Test accuracy
            value: 53.7
          - type: accuracy
            name: Breton Test accuracy
            value: 68.7
          - type: accuracy
            name: Telugu Test accuracy
            value: 82.9
          - type: accuracy
            name: Cantonese Test accuracy
            value: 57
          - type: accuracy
            name: Old Church Slavonic Test accuracy
            value: 59.1
          - type: accuracy
            name: Karelian Test accuracy
            value: 75
          - type: accuracy
            name: Upper Sorbian Test accuracy
            value: 77.8
          - type: accuracy
            name: South Levantine Arabic Test accuracy
            value: 71.2
          - type: accuracy
            name: Komi Zyrian Test accuracy
            value: 47
          - type: accuracy
            name: Irish Test accuracy
            value: 69.4
          - type: accuracy
            name: Nayini Test accuracy
            value: 56.4
          - type: accuracy
            name: Munduruku Test accuracy
            value: 29.2
          - type: accuracy
            name: Manx Test accuracy
            value: 38.8
          - type: accuracy
            name: Skolt Sami Test accuracy
            value: 43.7
          - type: accuracy
            name: Afrikaans Test accuracy
            value: 88.2
          - type: accuracy
            name: Old Turkish Test accuracy
            value: 37.1
          - type: accuracy
            name: Tupinamba Test accuracy
            value: 44.5
          - type: accuracy
            name: Belarusian Test accuracy
            value: 90.4
          - type: accuracy
            name: Serbian Test accuracy
            value: 89.5
          - type: accuracy
            name: Moksha Test accuracy
            value: 49.1
          - type: accuracy
            name: Western Armenian Test accuracy
            value: 82
          - type: accuracy
            name: Scottish Gaelic Test accuracy
            value: 63.1
          - type: accuracy
            name: Khunsari Test accuracy
            value: 47.3
          - type: accuracy
            name: Hebrew Test accuracy
            value: 88.5
          - type: accuracy
            name: Uyghur Test accuracy
            value: 78
          - type: accuracy
            name: Chukchi Test accuracy
            value: 37.5

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

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