proycon/colibri-core

Issues

• Class encoding fails if input only contains one line without new line?

  #34 opened 8 years ago by proycon
  1

• Package for Alpine Linux

  #45 opened 3 years ago by proycon
  1

• Implement begin-of-sentence and end-of-sentence markers in pattern model training

  #11 opened 9 years ago by proycon
  1

• Can't compile on CentOS 6.6

  #22 opened 5 years ago by andreasvc
  2

• Wrong threshold in model.filter

  #39 opened 5 years ago by svetlana21
  3

• Non-functioning constraints in .getrightneighbours(), .getcooc() etc.

  #44 opened 5 years ago by KasperFyhn
  2

• Problems compiling with anaconda

  #43 opened 6 years ago by redadmiral
  1

• Unable to load large corpora into memory because PatternPointer length can't exceed 2^32 bytes (32 bit size descriptor)

  #42 opened 6 years ago by proycon
  5

• Refactor Pattern.instanceof to make use of v2 changes and add flexgram support

  #18 opened 9 years ago by proycon
  1

• Missing data in indexed model on large data set; yields much lower counts than unindexed model on the same data with the same parameters!

  #41 opened 6 years ago by proycon
  4

• how to expose colibri-ngrams from Python API?

  #38 opened 7 years ago by mikkokotila
  4

• Error with Tibetan Unicode

  #37 opened 7 years ago by ngawangtrinley
  2

• [Queries] Ability to create a model and cls from multiple input files

  #36 opened 7 years ago by manrock007
  0

• Implement ability to filter on (n)PMI for getleftneighbours(), getleftcooc(), etc..

  #33 opened 8 years ago by proycon
  0

• Clean warnings in v2

  #3 opened 8 years ago by proycon
  3

• Investigate improved scalability using use of out-of-memory datastructures

  #27 opened 8 years ago by proycon
  1

• tokens/coverage results not split out per n category?

  #26 opened 8 years ago by proycon
  1

• buildpattern() does not raise an exception when unknown tokens are presented in the input and allowunknown=false (default)!

  #25 opened 8 years ago by proycon
  0

• Provide vocabulary file

  #2 opened 9 years ago by naiaden
  1

• Process comments of reviewers of the Colibri Core paper

  #24 opened 9 years ago by proycon
  3

• Load corpora with mmap

  #23 opened 9 years ago by andreasvc
  1

• Discrepancy between totaloccurrencesingroup and patterns in getreverseindex

  #20 opened 9 years ago by naiaden
  3

• Pattern.ngrams() performance too slow for very large patterns, can be sped up

  #21 opened 9 years ago by proycon
  0

• pip failed building wheel for colibricore Mac OSX 10.11.2

  #12 opened 9 years ago by mkrump
  4

• No flexgram support in IndexedPatternModel.getsubchildren() / getsubparents() yet

  #19 opened 9 years ago by proycon
  0

• Support initial and final skips in skipgram training? [vote if desired]

  #8 opened 9 years ago by proycon
  0

• PatternModel.getreverseindex() is too slow with skipgrams

  #17 opened 9 years ago by proycon
  1

• skipgram training adds strange ngrams that do not exist

  #16 opened 9 years ago by proycon
  2

• lower-order ngrams not pruned when training with skipgrams, minlengh > 1 and t > 1

  #15 opened 9 years ago by proycon
  0

• reverse() is broken, doesn't work for higher classes (>128)

  #14 opened 9 years ago by proycon
  1

• Implement more efficient algorithms for the search and extraction of pre-specified skipgrams and flexgrams

  #9 opened 9 years ago by proycon
  0

• getskipcontent() broken in v2

  #10 opened 9 years ago by proycon
  1

• Python Tutorial needs an update for v2

  #6 opened 9 years ago by proycon
  1

• iterating over model.reverseindex() crashes in Python

  #5 opened 9 years ago by proycon
  1

• Python: catch C++ exception when training indexed model without reverseindex

  #7 opened 9 years ago by proycon
  0

• Check in advance for reverse index when training skipgrams with IndexedPatternModel

  #4 opened 9 years ago by proycon
  0

• computestats() is too slow, delaying saving of indexed model significantly

  #1 opened 9 years ago by proycon
  1