Biaffine-based UD Parser 22.10

Jasonarson, Atli; Steingrímsson, Steinþór; Sigurðsson, Einar Freyr; Daðason, Jón Friðrik

Biaffine-based UD Parser 22.10

Clarin IS Repository

Authors: Jasonarson, Atli ; Steingrímsson, Steinþór ; Sigurðsson, Einar Freyr and Daðason, Jón Friðrik

Item identifier: http://hdl.handle.net/20.500.12537/273

Date issued: 2022-10-01

Type: toolService

Language(s): Icelandic

Description: ENGLISH: This Universal Dependencies parser for Icelandic was trained with Diaparser [1] on IcePaHC [2] and UD_Icelandic-Modern [3], the latter one having been revised before training, as some duplicate sentences had to be removed. The parser utilizes information from an ELECTRA language model [4]. Its UAS (unlabeled attachment score) is 89.52 and its LAS (labeled attachment score) is 86.23.

ICELANDIC: Þessi UD-þáttari var þjálfaður með Diaparser [1] á IcePaHC [2] og UD_Icelandic-Modern [3] en síðarnefnda málheildin var uppfærð fyrir þjálfun tólsins, þar sem fjarlægðar voru úr henni endurteknar setningar. Þáttarinn nýtir sér upplýsingar úr ELECTRA-mállíkani [4]. Hann skorar 89.52 á UAS (unlabeled attachment score) og 86.23 á LAS (labeled attachment score). [1] Diaparser: https://github.com/Unipisa/diaparser [2] IcePaHC: https://github.com/UniversalDependencies/UD_Icelandic-IcePaHC/ [3] UD_Icelandic-Modern: https://github.com/UniversalDependencies/UD_Icelandic-Modern/ [4] electra-base-igc-is: https://huggingface.co/jonfd/electra-base-igc-is

Publisher: The Árni Magnússon Institute for Icelandic Studies

Subject(s): universal dependencies parsing

Collection(s): Clarin IS

This item is replaced by a newer submission:

http://hdl.handle.net/20.500.12537/302

Show full item record

Files in this item

Download all files in item (831.59 MB)

This item is

Publicly Available

and licensed under:
Apache License 2.0

Name: biaffine-dependency-parser.zip
Size: 831.58 MB
Format: application/zip
Description: Biaffine-dependency-parser
MD5: 467f558f6aa6b7e5a6b36ebddff43d57

Download file Preview

File Preview

biaffine-dependency-parser
- transformer_models
  - electra-base-igc-is
    - config.json466 B
    - README.md666 B
    - tokenizer_config.json73 B
    - special_tokens_map.json112 B
    - pytorch_model.bin422 MB
    - .gitattributes1 kB
    - vocab.txt253 kB
- parse_file.py607 B
- Tokenizer
  - src
    - tokenizer
      - __init__.py2 kB
      - tokenizer.py125 kB
      - __pycache__
        abbrev.cpython-38.pyc8 kB
        __init__.cpython-38.pyc2 kB
        tokenizer.cpython-38.pyc62 kB
        definitions.cpython-38.pyc15 kB
        version.cpython-38.pyc194 B
      - abbrev.py13 kB
      - definitions.py27 kB
      - version.py22 B
      - py.typed0 B
      - main.py9 kB
      - Abbrev.conf46 kB
  - setup.py3 kB
  - .gitignore1 kB
  - README.rst39 kB
  - setup.cfg64 B
  - test
    - toktest_large.txt559 kB
    - test_tokenizer_tok.py18 kB
    - toktest_normal.txt12 kB
    - test_index_calculation.py20 kB
    - toktest_normal_gold_expected.txt13 kB
    - toktest_edgecases.txt6 kB
    - toktest_edgecases_gold_expected.txt6 kB
    - test_detokenize.py2 kB
    - Overview.txt34 kB
    - toktest_large_gold_perfect.txt576 kB
    - toktest_large_gold_acceptable.txt583 kB
    - example.txt3 kB
    - toktest_sentences.txt21 kB
    - toktest_edgecases_diff.txt751 B
    - test_tokenizer.py102 kB
  - .github
    - workflows
      - python-package.yml1 kB
  - .git
    - logs
      - refs
        remotes
        origin
        HEAD176 B
        heads
        master176 B
      - HEAD176 B
    - info
      - exclude240 B
    - config261 B
    - packed-refs2 kB
    - index2 kB
    - HEAD23 B
    - refs
      - remotes
        origin
        HEAD32 B
      - tags
      - heads
        master41 B
    - description73 B
    - hooks
      - push-to-checkout.sample2 kB
      - applypatch-msg.sample478 B
      - commit-msg.sample896 B
      - pre-push.sample1 kB
      - pre-rebase.sample4 kB
      - post-update.sample189 B
      - pre-receive.sample544 B
      - pre-applypatch.sample424 B
      - update.sample3 kB
      - pre-commit.sample1 kB
      - pre-merge-commit.sample416 B
      - fsmonitor-watchman.sample4 kB
      - prepare-commit-msg.sample1 kB
    - objects
      - pack
        pack-258487a22aa4bdca999801128b1c799f60c4d907.idx74 kB
        pack-258487a22aa4bdca999801128b1c799f60c4d907.pack988 kB
      - info
    - branches
  - release.sh251 B
  - LICENSE1 kB
  - MANIFEST.in74 B
- electra-ud-parser
  - parser.train.log99 kB
  - parser474 MB
- requirements.txt36 B
- test_file.txt103 B

Name: README.md
Size: 1.97 KB
Format: Unknown
Description: Readme
MD5: bffde57167f046cffc8bd2bf8b1e5552

Download file

Biaffine-based UD Parser 22.10

Files in this item

Partners, Coordination, Funding

Repository

More