Commits · 9ff007c0366495b37ee330e969d661c01128003d · Franck Dary / macaon_data

Aug 06, 2020
- Updated machines and added regression machine timer · 9ff007c0
  Franck Dary authored 4 years ago
  
  9ff007c0
Jul 28, 2020
- Produce pretrained glove embeddings for multiple columns and for letters · 42de4d6d
  Franck Dary authored 4 years ago
  
  42de4d6d
- Using GloVe by default to pretrain word embeddings · 70f6b20a
  Franck Dary authored 4 years ago
  
  70f6b20a
Jul 07, 2020
- Changed some transitions · 3e832ec8
  Franck Dary authored 4 years ago
  
  3e832ec8
Jul 03, 2020
- Give more transitions to tokenizer · b74a8e24
  Franck Dary authored 4 years ago
  
  b74a8e24
Jul 02, 2020
- Updated tokenizer · 384b3fb1
  Franck Dary authored 4 years ago
  
  384b3fb1
Jun 13, 2020
- Using w2v on output from udpipe horizontal output instead of on raw text · 18f1683c
  Franck Dary authored 4 years ago
  
  18f1683c
Jun 10, 2020
- Corpora are now copied into exp dir to avoid conflicts between multiple experiences · 26f25b22
  Franck Dary authored 4 years ago
  
  26f25b22
Jun 09, 2020
- Added support for lemmatization · a301e461
  Franck Dary authored 4 years ago
  
  a301e461
Jun 03, 2020
- scripts now use conllu mcd by default, and can read mcd from file metadata · 4369a755
  Franck Dary authored 4 years ago
  
  4369a755
May 17, 2020
- Added script to pretrain embeddings on w2v · b17524f7
  Franck Dary authored 4 years ago
  
  b17524f7
May 03, 2020
- updated tagparser_seq for new strategy · 6607f37a
  Franck Dary authored 4 years ago
  
  6607f37a
Apr 28, 2020
- Removed l_rules from Makefile · e73391bf
  Franck Dary authored 4 years ago
  
  e73391bf
Apr 06, 2020
- take dev into account when creating splittransitions · 69278bd4
  Franck Dary authored 5 years ago
  
  69278bd4
Apr 03, 2020
- Added more SPLIT transitions because arabic needs it · 07795c9d
  Franck Dary authored 5 years ago
  
  07795c9d
Apr 02, 2020
- Adapted to new split transition · ac4b5d0a
  Franck Dary authored 5 years ago
  
  ac4b5d0a
Mar 31, 2020
- Changed the way we train and evaluate · 9074e66f
  Franck Dary authored 5 years ago
  
  9074e66f
Mar 21, 2020
- action IGNORECHAR now has lowest priority · 4f9f249e
  Franck Dary authored 5 years ago
  
  4f9f249e
Mar 07, 2020
- Updated for tokenizer · f1f18db2
  Franck Dary authored 5 years ago
  
  f1f18db2
Mar 05, 2020
- Added tokenizer machine · 4ed5c4e0
  Franck Dary authored 5 years ago
  
  4ed5c4e0
Feb 27, 2020
- Made the transitionsSet generation script to output expected state name · 901b4ab2
  Franck Dary authored 5 years ago
  
  901b4ab2
Feb 17, 2020
- Removed Default qualifier for transitions · b9dd9c0b
  Franck Dary authored 5 years ago
  
  b9dd9c0b
Jan 06, 2020
- Initial commit · c00fe2d1
  Franck Dary authored 5 years ago
  
  c00fe2d1