DeNederlandscheBank/nqm

Issues

• Make it DNB ready

  #68 opened 3 years ago
  0

• Evluation failed for subwords models: E, F, I, J

  #67 opened 3 years ago
  0

• Business Documentatie (30-06-2021)

  #64 opened 3 years ago
  0

• Project finalization

  #63 opened 3 years ago
  0

• Update README of project

  #62 opened 3 years ago
  0

• Adapt Evaluation

  #56 opened 4 years ago
  1

• Script folder

  #55 opened 4 years ago
  0

• Change use of copy script

  #54 opened 4 years ago
  1

• enhance fairseq-preprocess for several test files

  #53 opened 4 years ago
  1

• Folder XLMR.base adaptions

  #52 opened 4 years ago
  0

• Housekeeping work for new model run

  #51 opened 3 years ago
  0

• Check results on Test 2 for model GOLF

  #50 opened 4 years ago
  1

• Subwords/BPE models have unk in the inputs

  #49 opened 3 years ago
  2

• Experiment with One-Shot Translation

  #48 opened 3 years ago
  1

• quotation marks in names cause struggle in identifyingName

  #45 opened 4 years ago
  0

• move subword_nmt to root

  #44 opened 4 years ago
  0

• EXAMPLES_PER_EXAMPLE is not working

  #41 opened 4 years ago
  0

• Evaluation for pg-model not working correctly

  #40 opened 4 years ago
  0

• Pointer-generator architecture to improve name translation

  #39 opened 4 years ago
  0

• evaluation not correctly working

  #38 opened 4 years ago
  0

• fast_align cannot be updated on new machine

  #37 opened 4 years ago
  0

• Wrong templates used for training models --> retrain all models

  #36 opened 4 years ago
  0

• fairseq (0.10.2) replace-unk broken because of wrong data format

  #35 opened 4 years ago
  0

• fairseq-generate not working with replace-unk

  #33 opened 4 years ago
  1

• Placeholder mechanism for name translation

  #32 opened 4 years ago
  2

• Copy mechanism for names

  #31 opened 4 years ago
  1

• Add accuracy to evaluation

  #30 opened 4 years ago
  1

• Bilingual Model for dutch and english questions

  #29 opened 3 years ago
  1

• Use pre-trained model

  #26 opened 3 years ago
  9

• More templates are required

  #25 opened 4 years ago
  1

• Implement name check functionality

  #24 opened 4 years ago
  1

• Adapt templates

  #23 opened 4 years ago
  0

• Evaluation on multiple test sets

  #22 opened 4 years ago
  2

• EIOPA Dataset Generator

  #19 opened 4 years ago
  1

• Add large english dictionary

  #18 opened 4 years ago
  1

• Solution for unknown names

  #17 opened 4 years ago
  1

• [High priority] Data generation, tokenization and BPE missing

  #16 opened 4 years ago
  1

• Names and/or IDs not correctly translated

  #15 opened 4 years ago
  2

• Add Model Evaluation Process

  #14 opened 4 years ago
  2

• Large file storage

  #13 opened 3 years ago
  3

• Bug when fairseq training is ended

  #10 opened 4 years ago
  2

• Generation of test sets is slow

  #9 opened 4 years ago
  0

• EIOPA Generator

  #8 opened 4 years ago
  0

• Adaptions to Database

  #7 opened 4 years ago
  3

• URI's not human readable

  #6 opened 4 years ago
  0

• Directory structure

  #5 opened 4 years ago
  0

• Things to change in generator for different dataset

  #4 opened 4 years ago
  1

• No saving of models and datasets

  #2 opened 4 years ago
  1

• Existing Pipeline broken

  #1 opened 4 years ago
  0