Commits

Branch selector

 acl2018

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 25, 2018

• Release v0.0.3

  nikitakit

  committed
  57d9043

  Copy full SHA for 57d9043

• Throw error when a sentence is too long

  Show description for 536bd3e
  nikitakit

  committed
  536bd3e

  Copy full SHA for 536bd3e

Commits on Aug 6, 2018

• Add parse command to parse raw sentences using pytorch code

  nikitakit

  committed
  bc23dd2

  Copy full SHA for bc23dd2

Commits on Jun 20, 2018

• Release v0.0.2

  nikitakit

  committed
  18f5792

  Copy full SHA for 18f5792

Commits on Jun 19, 2018

• Fix IndexError due to bad bounds check in spaCy plugin

  nikitakit

  committed
  b9f8359

  Copy full SHA for b9f8359

• Update installation to better enforce tensorflow compatibility

  Show description for f3b6423
  nikitakit

  committed
  f3b6423

  Copy full SHA for f3b6423

• Attempt to fix locating numpy headers on Mac built-in python

  Show description for c120086
  nikitakit

  committed
  c120086

  Copy full SHA for c120086

Commits on Jun 10, 2018

• Release more English parsing models

  nikitakit

  committed
  5567867

  Copy full SHA for 5567867

• Include tensorflow version requirement in readme

  nikitakit

  committed
  ba6cc86

  Copy full SHA for ba6cc86

Commits on May 29, 2018

• Parser release for inference

  nikitakit

  committed
  bf9df44

  Copy full SHA for bf9df44

• Add ELMo command to EXPERIMENTS.md

  nikitakit

  committed
  63c3095

  Copy full SHA for 63c3095

Commits on Apr 27, 2018

• Consolidated code for self-attentive model

  Show description for bf5bbae
  nikitakit

  committed
  bf5bbae

  Copy full SHA for bf5bbae

Commits on Dec 5, 2017

• Use cython for chart decode

  nikitakit

  committed
  f5ca5be

  Copy full SHA for f5ca5be

Commits on Dec 4, 2017

• Remove post-lstm dropout

  nikitakit

  committed
  a5a7224

  Copy full SHA for a5a7224

Commits on Dec 2, 2017

• Add more dropout around the LSTM

  nikitakit

  committed
  3842f88

  Copy full SHA for 3842f88

• Use CUDA with pytorch

  nikitakit

  committed
  36904c6

  Copy full SHA for 36904c6

Commits on Dec 1, 2017

• Fix no-loss handling

  nikitakit

  committed
  c326761

  Copy full SHA for c326761

• Fix returning non-Variable loss

  nikitakit

  committed
  70f9aa8

  Copy full SHA for 70f9aa8

• Fix switching between pytorch and dynet

  nikitakit

  committed
  8541321

  Copy full SHA for 8541321

• Use pytorch dropout

  nikitakit

  committed
  5caeefe

  Copy full SHA for 5caeefe

• Implement saving when using pytorch

  nikitakit

  committed
  9442b8e

  Copy full SHA for 9442b8e

• Allow running even if only one of pytorch or dynet is available

  nikitakit

  committed
  86bff9d

  Copy full SHA for 86bff9d

• Remember to zero grad for pytorch

  nikitakit

  committed
  43e22ca

  Copy full SHA for 43e22ca

• Speed up pytorch code by precomputing all label scores

  Show description for 64ed13b
  nikitakit

  committed
  64ed13b

  Copy full SHA for 64ed13b

• Initial pytorch port

  nikitakit

  committed
  01cc97a

  Copy full SHA for 01cc97a

Commits on Nov 15, 2017

• Added a clarification to the README

  mitchellstern

  authored
  ff685cd

  Copy full SHA for ff685cd

• Renamed oracle_split method to oracle_splits

  mitchellstern

  committed
  1f3e638

  Copy full SHA for 1f3e638

• Updated the wording in the README

  mitchellstern

  authored
  eb20df7

  Copy full SHA for eb20df7

• Updated README to include information about chart parser

  mitchellstern

  authored
  e15a4c3

  Copy full SHA for e15a4c3

• Added a pre-trained chart model

  mitchellstern

  committed
  a839654

  Copy full SHA for a839654

• Added code for chart parser

  mitchellstern

  committed
  e4fd8fd

  Copy full SHA for e4fd8fd

Commits on Nov 4, 2017

• Added default paths for EVALB and data arguments

  mitchellstern

  committed
  44b387f

  Copy full SHA for 44b387f

• Added pre-trained model extraction instructions to setup

  mitchellstern

  committed
  e9c80fc

  Copy full SHA for e9c80fc

• Updated .gitignore to ignore uncompressed models

  mitchellstern

  committed
  9115616

  Copy full SHA for 9115616

• Added a pre-trained top-down model

  mitchellstern

  committed
  3c0c86a

  Copy full SHA for 3c0c86a

Pagination

Previous

Next