Commits

Branch selector

 master

User selector

AngledLuffa

Datepicker

All time

Commit History

Commits on Jan 9, 2015

• Set default port for the NERServer to 4465

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  a6890f2

  Copy full SHA for a6890f2

Commits on Jan 8, 2015

• Add a transform which puts the sentence final punctuation outside the rest of the tree structure (a hack, in other words)

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  6b8e4ad

  Copy full SHA for 6b8e4ad

Commits on Dec 20, 2014

• hmmm testing?

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  d9a4440

  Copy full SHA for d9a4440

Commits on Dec 18, 2014

• Update parser training scripts for the new home of the handwritten trees

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  94bc69e

  Copy full SHA for 94bc69e

Commits on Dec 12, 2014

• Merge branch 'master' of /u/nlp/git/javanlp

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  b5717c7

  Copy full SHA for b5717c7

• Make a couple member variables private

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  1e9d390

  Copy full SHA for 1e9d390

Commits on Dec 11, 2014

• Merge branch 'master' of /u/nlp/git/javanlp

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  4930211

  Copy full SHA for 4930211

• Undo a broken part of the recent CoreLabel change

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  b324ae7

  Copy full SHA for b324ae7

Commits on Dec 10, 2014

• User submitted efficiency improvement: only initialize this set once. From Paolo Dragone

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  462c0cd

  Copy full SHA for 462c0cd

Commits on Dec 8, 2014

• Add a link in javadoc

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  3336325

  Copy full SHA for 3336325

• Add a method for getting the known states from a model

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  8a1ec63

  Copy full SHA for 8a1ec63

Commits on Dec 4, 2014

• Merge branch 'master' of /u/nlp/git/javanlp

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  c8a1a96

  Copy full SHA for c8a1a96

• Add some more notes on how to run this

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  0f76293

  Copy full SHA for 0f76293

Commits on Dec 2, 2014

• Test immediately after an epoch, rather than waiting for the start of the next epoch

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  0f678a1

  Copy full SHA for 0f678a1

• Use more precision when writing out matrices

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  d9bb2c3

  Copy full SHA for d9bb2c3

• Add an option to write out the initial matrices

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  bee7171

  Copy full SHA for bee7171

• Add an option which turns off shuffling matrices

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  ac0ec34

  Copy full SHA for ac0ec34

Commits on Nov 19, 2014

• Add code for outputting everything at once to the sentiment model

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  7ab2684

  Copy full SHA for 7ab2684

• Apply the CoordinationTransformer after making the first copy for the EnglishGrammaticalStructure. Saves time by only copying once, and the CoordinationTransformer is a lot faster with the TreeGrap…

  Show description for 197ebb2
  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  197ebb2

  Copy full SHA for 197ebb2

• Override add/remove child, fix setchildren so that it maintains parent pointers when the TreeGraphNode nodes are manipulated

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  b0a87d6

  Copy full SHA for b0a87d6

• When making new nodes in tsurgeon, you need to make nodes which match the tree you are matching, not the tree which is storing the tsurgeon pattern

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  765dc2b

  Copy full SHA for 765dc2b

• A couple minor initialization and regularization changes to make the java sentiment code more like the matlab sentiment code

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  90f7522

  Copy full SHA for 90f7522

• Dependency graphs should now be able to safely reuse the leaves they are created with. Fixes some unexpected behavior in corenlp

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  6905952

  Copy full SHA for 6905952

Commits on Nov 18, 2014

• Remove HeadWordAnnotation. The one place that actually cares about it, the dependency converter, now uses something equivalent attached directly on the TreeGraphNode

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  57e60a4

  Copy full SHA for 57e60a4

• Remove the links back from TreeGraphNode to GrammaticalStructure

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  aafd992

  Copy full SHA for aafd992

• Remove unused constructor

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  4e737ab

  Copy full SHA for 4e737ab

• Merge branch 'master' of /u/nlp/git/javanlp

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  56382e8

  Copy full SHA for 56382e8

• Remove ProjectiveCategory, which is not used anywhere

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  d7d6fdf

  Copy full SHA for d7d6fdf

• Remove all uses of HeadTagAnnotation

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  0338040

  Copy full SHA for 0338040

• Change the srparser to use pointers to labels instead of words for head words & head tags

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  42d7b2b

  Copy full SHA for 42d7b2b

• No one cares about HeadTagNode (srparser does, but it sets HeadTagNode itself)

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  b125712

  Copy full SHA for b125712

Commits on Nov 17, 2014

• Only regularize words which are actively used

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  b085431

  Copy full SHA for b085431

• Add some doc

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  f3c965e

  Copy full SHA for f3c965e

• Add a check of the regularization component of the sentiment gradient. This is a much smaller factor, so its component needs to be turned up a lot to verify it is working

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  fe8e9a4

  Copy full SHA for fe8e9a4

Commits on Nov 15, 2014

• Merge branch 'master' of /u/nlp/git/javanlp

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  3e00b1e

  Copy full SHA for 3e00b1e

Pagination

Previous

Next