Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 4, 2024

• Add a prop file for a newer version of the French WikiNER dataset

  AngledLuffa

  committed
  698d84b

  Copy full SHA for 698d84b

Commits on Oct 30, 2021

• Save the multiword file for Italian and then use it at runtime. Another alternative would have been to save it directly into the model

  AngledLuffa

  committed
  5d33ab4

  Copy full SHA for 5d33ab4

Commits on Sep 30, 2021

• Add blocks for building Italian SR parsers from the Turin datasets

  AngledLuffa

  committed
  239633b

  Copy full SHA for 239633b

• Add skeletal tlp and tlpp for Italian

  AngledLuffa

  committed
  4d0305f

  Copy full SHA for 4d0305f

• Add support for Italian tokenizer, NER, depparse, and tagger

  AngledLuffa

  committed
  2740c30

  Copy full SHA for 2740c30

• Add a block to create a Hungarian constituency parser

  AngledLuffa

  committed
  81c1b71

  Copy full SHA for 81c1b71

• Add a Hungarian NER model

  AngledLuffa

  committed
  68f72aa

  Copy full SHA for 68f72aa

• Add a Hungarian tagger built from UD

  AngledLuffa

  committed
  cfe8ee1

  Copy full SHA for cfe8ee1

• Add a Makefile for creating a Hungarian stattok

  AngledLuffa

  committed
  6972cd8

  Copy full SHA for 6972cd8

• Update tagger Makefile with the latest German UD model

  AngledLuffa

  committed
  39635aa

  Copy full SHA for 39635aa

Commits on Mar 16, 2021

• Merge pull request #1142 from stanfordnlp/sr_custom_serialization

  Show description for 83426ab
  AngledLuffa

  authored
  83426ab

  Copy full SHA for 83426ab

• Apparently the latest modifications use even more memory

  AngledLuffa

  committed
  7e9eaf1

  Copy full SHA for 7e9eaf1

Commits on Mar 7, 2021

• Increase scripts/lexparser/lexparser.sh mem usage to 200m

  Show description for 43a2d2b
  mattmoore

  committed
  43a2d2b

  Copy full SHA for 43a2d2b

Commits on Feb 11, 2021

• The table of transitions looks like a useful bit of statistics to put at the end of a test run

  AngledLuffa

  committed
  6978f9e

  Copy full SHA for 6978f9e

• Add -quietEvaluation to the test phase of the SRParser

  AngledLuffa

  committed
  023514a

  Copy full SHA for 023514a

Commits on Jan 27, 2021

• Add an argument for augmenting fewer sentences. Also, change the name to something more understandable

  AngledLuffa

  committed
  d8ce461

  Copy full SHA for d8ce461

Commits on Jan 22, 2021

• Chinese shards 5... increases performance by a whopping 0.2%, only takes an extra two days to train

  AngledLuffa

  committed
  859d098

  Copy full SHA for 859d098

Commits on Dec 29, 2020

• Add a comment on how to make a GSD segmenter

  AngledLuffa

  committed
  2df7c92

  Copy full SHA for 2df7c92

• Add a command line for a GSD segmenter

  AngledLuffa

  committed
  eee6259

  Copy full SHA for eee6259

• Add a line for testing the ctb9 segmenters

  AngledLuffa

  committed
  a0ff942

  Copy full SHA for a0ff942

• Make it possible to run CTB9 segmenter on my desktop

  AngledLuffa

  committed
  33386ee

  Copy full SHA for 33386ee

• Update comments -> echo so it stops being pink in emacs

  AngledLuffa

  committed
  2ec0e3b

  Copy full SHA for 2ec0e3b

Commits on Dec 16, 2020

• Retrain French with multiple shards

  AngledLuffa

  committed
  8521008

  Copy full SHA for 8521008

Commits on Nov 30, 2020

• Separate out two make targets

  AngledLuffa

  committed
  d1a2f28

  Copy full SHA for d1a2f28

• Unify & simplify options

  Show description for c60c7f6
  AngledLuffa

  committed
  c60c7f6

  Copy full SHA for c60c7f6

• Train WSJ-only models as well

  AngledLuffa

  committed
  2420fe4

  Copy full SHA for 2420fe4

• Update with some parameters that seem to produce a pretty good non-beam model

  AngledLuffa

  committed
  4c90eea

  Copy full SHA for 4c90eea

• Unify and simplify srparser flags a bit

  AngledLuffa

  committed
  4176b76

  Copy full SHA for 4176b76

Commits on Nov 21, 2020

• Update Makefile to work on other filesystems, especially my own, for the Arabic, Chinese, English, French, German, and Spanish sr parsers

  AngledLuffa

  committed
  9d2bd66

  Copy full SHA for 9d2bd66

Commits on Nov 20, 2020

• I bet that got good results!

  AngledLuffa

  committed
  817f621

  Copy full SHA for 817f621

Commits on Jul 17, 2020

• This value builds a model of comparable size to the old ctb7 models

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  6f5e1ce

  Copy full SHA for 6f5e1ce

• Oops, a typo

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  f288b3a

  Copy full SHA for f288b3a

• Add a smaller segmenter model to the chinese downloads

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  b36b469

  Copy full SHA for b36b469

• More memory needed for training the segmenters with lots of threads

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  03749e7

  Copy full SHA for 03749e7

• Add a ctb9 model with a smaller feature set

  

  AngledLuffa

  authored and

  Stanford NLP

  committed
  b596f17

  Copy full SHA for b596f17

Pagination

Previous

Next