wietsedv's picture
update README
32633d4
metadata
language:
  - uk
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-uk
    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: 82.2
          - type: accuracy
            name: Dutch Test accuracy
            value: 84.3
          - type: accuracy
            name: German Test accuracy
            value: 82.4
          - type: accuracy
            name: Italian Test accuracy
            value: 83.9
          - type: accuracy
            name: French Test accuracy
            value: 82.6
          - type: accuracy
            name: Spanish Test accuracy
            value: 86.2
          - type: accuracy
            name: Russian Test accuracy
            value: 93.3
          - type: accuracy
            name: Swedish Test accuracy
            value: 86.3
          - type: accuracy
            name: Norwegian Test accuracy
            value: 80.2
          - type: accuracy
            name: Danish Test accuracy
            value: 85.2
          - type: accuracy
            name: Low Saxon Test accuracy
            value: 30.9
          - type: accuracy
            name: Akkadian Test accuracy
            value: 17.5
          - type: accuracy
            name: Armenian Test accuracy
            value: 87.7
          - type: accuracy
            name: Welsh Test accuracy
            value: 66.8
          - type: accuracy
            name: Old East Slavic Test accuracy
            value: 77.5
          - type: accuracy
            name: Albanian Test accuracy
            value: 79.7
          - type: accuracy
            name: Slovenian Test accuracy
            value: 84.5
          - type: accuracy
            name: Guajajara Test accuracy
            value: 14.6
          - type: accuracy
            name: Kurmanji Test accuracy
            value: 77
          - type: accuracy
            name: Turkish Test accuracy
            value: 76.3
          - type: accuracy
            name: Finnish Test accuracy
            value: 82.5
          - type: accuracy
            name: Indonesian Test accuracy
            value: 77
          - type: accuracy
            name: Ukrainian Test accuracy
            value: 98.2
          - type: accuracy
            name: Polish Test accuracy
            value: 91.8
          - type: accuracy
            name: Portuguese Test accuracy
            value: 84.1
          - type: accuracy
            name: Kazakh Test accuracy
            value: 81.8
          - type: accuracy
            name: Latin Test accuracy
            value: 77.9
          - type: accuracy
            name: Old French Test accuracy
            value: 26.9
          - type: accuracy
            name: Buryat Test accuracy
            value: 60.7
          - type: accuracy
            name: Kaapor Test accuracy
            value: 5.4
          - type: accuracy
            name: Korean Test accuracy
            value: 61.5
          - type: accuracy
            name: Estonian Test accuracy
            value: 84.4
          - type: accuracy
            name: Croatian Test accuracy
            value: 93.2
          - type: accuracy
            name: Gothic Test accuracy
            value: 3.7
          - type: accuracy
            name: Swiss German Test accuracy
            value: 35
          - type: accuracy
            name: Assyrian Test accuracy
            value: 14.6
          - type: accuracy
            name: North Sami Test accuracy
            value: 27
          - type: accuracy
            name: Naija Test accuracy
            value: 22.5
          - type: accuracy
            name: Latvian Test accuracy
            value: 88.9
          - type: accuracy
            name: Chinese Test accuracy
            value: 51.9
          - type: accuracy
            name: Tagalog Test accuracy
            value: 71.1
          - type: accuracy
            name: Bambara Test accuracy
            value: 18.7
          - type: accuracy
            name: Lithuanian Test accuracy
            value: 88.1
          - type: accuracy
            name: Galician Test accuracy
            value: 85.8
          - type: accuracy
            name: Vietnamese Test accuracy
            value: 66.3
          - type: accuracy
            name: Greek Test accuracy
            value: 85.9
          - type: accuracy
            name: Catalan Test accuracy
            value: 84
          - type: accuracy
            name: Czech Test accuracy
            value: 92.1
          - type: accuracy
            name: Erzya Test accuracy
            value: 49.4
          - type: accuracy
            name: Bhojpuri Test accuracy
            value: 51.8
          - type: accuracy
            name: Thai Test accuracy
            value: 63.3
          - type: accuracy
            name: Marathi Test accuracy
            value: 88.3
          - type: accuracy
            name: Basque Test accuracy
            value: 75.7
          - type: accuracy
            name: Slovak Test accuracy
            value: 91.8
          - type: accuracy
            name: Kiche Test accuracy
            value: 22.7
          - type: accuracy
            name: Yoruba Test accuracy
            value: 20
          - type: accuracy
            name: Warlpiri Test accuracy
            value: 32.4
          - type: accuracy
            name: Tamil Test accuracy
            value: 81.7
          - type: accuracy
            name: Maltese Test accuracy
            value: 16.6
          - type: accuracy
            name: Ancient Greek Test accuracy
            value: 63
          - type: accuracy
            name: Icelandic Test accuracy
            value: 81.4
          - type: accuracy
            name: Mbya Guarani Test accuracy
            value: 23.7
          - type: accuracy
            name: Urdu Test accuracy
            value: 64.1
          - type: accuracy
            name: Romanian Test accuracy
            value: 82.6
          - type: accuracy
            name: Persian Test accuracy
            value: 78.3
          - type: accuracy
            name: Apurina Test accuracy
            value: 24.8
          - type: accuracy
            name: Japanese Test accuracy
            value: 38
          - type: accuracy
            name: Hungarian Test accuracy
            value: 82.2
          - type: accuracy
            name: Hindi Test accuracy
            value: 68.3
          - type: accuracy
            name: Classical Chinese Test accuracy
            value: 36.6
          - type: accuracy
            name: Komi Permyak Test accuracy
            value: 46
          - type: accuracy
            name: Faroese Test accuracy
            value: 73.6
          - type: accuracy
            name: Sanskrit Test accuracy
            value: 13.9
          - type: accuracy
            name: Livvi Test accuracy
            value: 59.5
          - type: accuracy
            name: Arabic Test accuracy
            value: 82.1
          - type: accuracy
            name: Wolof Test accuracy
            value: 18.5
          - type: accuracy
            name: Bulgarian Test accuracy
            value: 91.1
          - type: accuracy
            name: Akuntsu Test accuracy
            value: 15.2
          - type: accuracy
            name: Makurap Test accuracy
            value: 2.1
          - type: accuracy
            name: Kangri Test accuracy
            value: 51.4
          - type: accuracy
            name: Breton Test accuracy
            value: 59.3
          - type: accuracy
            name: Telugu Test accuracy
            value: 84.3
          - type: accuracy
            name: Cantonese Test accuracy
            value: 53.8
          - type: accuracy
            name: Old Church Slavonic Test accuracy
            value: 48
          - type: accuracy
            name: Karelian Test accuracy
            value: 68.6
          - type: accuracy
            name: Upper Sorbian Test accuracy
            value: 71.7
          - type: accuracy
            name: South Levantine Arabic Test accuracy
            value: 68.9
          - type: accuracy
            name: Komi Zyrian Test accuracy
            value: 40.4
          - type: accuracy
            name: Irish Test accuracy
            value: 66.2
          - type: accuracy
            name: Nayini Test accuracy
            value: 46.2
          - type: accuracy
            name: Munduruku Test accuracy
            value: 8
          - type: accuracy
            name: Manx Test accuracy
            value: 23
          - type: accuracy
            name: Skolt Sami Test accuracy
            value: 27.7
          - type: accuracy
            name: Afrikaans Test accuracy
            value: 81.7
          - type: accuracy
            name: Old Turkish Test accuracy
            value: 39.8
          - type: accuracy
            name: Tupinamba Test accuracy
            value: 20.2
          - type: accuracy
            name: Belarusian Test accuracy
            value: 93.7
          - type: accuracy
            name: Serbian Test accuracy
            value: 93.8
          - type: accuracy
            name: Moksha Test accuracy
            value: 46
          - type: accuracy
            name: Western Armenian Test accuracy
            value: 79.8
          - type: accuracy
            name: Scottish Gaelic Test accuracy
            value: 56.3
          - type: accuracy
            name: Khunsari Test accuracy
            value: 36.5
          - type: accuracy
            name: Hebrew Test accuracy
            value: 84.4
          - type: accuracy
            name: Uyghur Test accuracy
            value: 77.2
          - type: accuracy
            name: Chukchi Test accuracy
            value: 35

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

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