nlp-uoregon/trankit

Issues

• Prevent from splitting on hyphen when doing tokenization for POS?

  #46 opened 3 years ago by olegpolivin
  1

• Trankit training for an unsupported language and with a custom embedding model

  #89 opened 3 months ago by Futyn-Maker
  0

• KeyError: 'lemma'

  #48 opened 3 years ago by Bachstelze
  2

• Cannot access http://nlp.uoregon.edu/download/trankit/v1.0.0/xlm-roberta-base/afrikaans.zip

  #88 opened 5 months ago by hungchuntse
  3

• MWT Training error

  #87 opened 7 months ago by nagaraju291990
  0

• lemmatize_doc ignores POS on pre-tokenized input.

  #86 opened 8 months ago by maayanorner
  0

• Not able to install trankit

  #81 opened 10 months ago by Lidan0241
  2

• Upgrade to Adapters library and recent Transformers versions

  #77 opened 9 months ago by calpt
  4

• trankit2conllu() does not write sentence texts: # text = ...

  #85 opened 9 months ago by jerekupari
  0

• Cuda device instead of use_gpu=True when multiple GPUs are available

  #84 opened 9 months ago by mrgransky
  0

• lemma with # sign in Finnish language

  #70 opened 2 years ago by mrgransky
  2

• OSError: Can't load weights for 'xlm-roberta-base'.

  #83 opened 10 months ago by xy1137030414
  4

• CUDA error: device-side assert triggered

  #52 opened 3 years ago by kkashleva
  2

• download_missing_files seems not to provide all missing files, in particular *_mwt_expander.pt (dutch)

  #80 opened a year ago by tcbrouwer
  1

• Training data with annotation scheme other than the UD one

  #79 opened a year ago by gcelano
  0

• How to get children from a particular token in trankit

  #63 opened a year ago by HitheshSankararaman
  1

• Does trankit support code-mixed languages (Two languages in one sentence)

  #64 opened a year ago by hithesh-sankararaman
  1

• OSError: Can't load config for 'xlm-roberta-base'.

  #65 opened a year ago by kirianguiller
  2

• Impossible to train on CPU

  #71 opened a year ago by n2oblife
  1

• Cannot train a new NER model on cpu-only machine

  #34 opened a year ago by oterrier
  2

• How to use my own word segmentation when doing part-of-speech tagging？

  #49 opened a year ago by CrazyCharles6
  1

• Hebrew link is broken

  #54 opened a year ago by talbaumel
  1

• Inconsistent tokenisation in French

  #68 opened a year ago by JakobMichiels
  2

• Trankit does not work on new versions of Python

  #73 opened a year ago by fergusq
  1

• While importing trankit getting mutable default error

  #76 opened a year ago by ArvinGoyal
  1

• Issue with lemmatization of indonesian

  #45 opened a year ago by xavier-taylor
  1

• switch base model from xlm-roberta-base

  #41 opened a year ago by Sneriko
  2

• Parse error of Italian

  #40 opened a year ago by gifdog97
  1

• with version 1.1.0 I get ModuleNotFoundError: No module named 'transformers'

  #37 opened a year ago by lfoppiano
  3

• Single shared configuration for all Pipeline instances

  #36 opened a year ago by abcdenis
  1

• Add a new pretrained model for Arabic...

  #31 opened a year ago by oterrier
  1

• Training Slovenian (and possibly other) customized lemmatize models produces incorrect predictions with <UNK> signs

  #75 opened a year ago by lkrsnik
  0

• Is it possible to extend Trankit for target language generation?

  #72 opened 2 years ago by yash-srivastava19
  3

• Can I use my only pretrained 'xlm-roberta-base' model?

  #69 opened 2 years ago by AIikai
  0

• Memory leaks, CUDA errors, and the state of this project

  #67 opened 2 years ago by lemontheme
  0

• UnicodeDecodeError: 'charmap' codec can't decode byte 0x81 in position 223: character maps to <undefined>

  #66 opened 2 years ago by MustafaCeyhan
  0

• GPU on Apple M1 chip support

  #62 opened 2 years ago by carschno
  0

• Error in tesing phase after training and creating a pipeline (posdep)

  #60 opened 2 years ago by Jeevesh28
  0

• Query about Java Based compatibility or ONNX model availability of trankit sentence segmentation module

  #59 opened 2 years ago by janmejay03
  0

• Memory leak in Pipeline() on a CPU

  #51 opened 3 years ago by navotoz
  3

• Finetuning Trankit for POS on a new language

  #35 opened 2 years ago by olegpolivin
  2

• Limit input string to 512 characters to avoid CUDA crash

  #58 opened 2 years ago by ulf1
  1

• RuntimeError: CUDA error: no kernel image is available for execution on the device

  #50 opened 3 years ago by baiziyuandyufei
  2

• Hardware recommendations

  #57 opened 2 years ago by hobodrifterdavid
  0

• How to load my custom pretrained dependency parser trained using Trankit for inference?

  #55 opened 2 years ago by Jivnesh
  1

• Output probabilities

  #53 opened 3 years ago by amir-zeldes
  0

• Add adapters for more model types

  #47 opened 3 years ago by Bachstelze
  0

• ImportError: cannot import name 'Pipeline' from partially initialized module 'trankit'

  #44 opened 3 years ago by EagleEyeKestrel
  1

• Torch version issue

  #33 opened 3 years ago by kpister
  3

• em dash character crashes French pipeline

  #42 opened 3 years ago by pa-nlp
  0