Code

Copy update#172

Code

Open

pskrunner14 wants to merge 80 commits intoIBM:copyIBM/pytorch-seq2seq:copyfrom pskrunner14:copypskrunner14/pytorch-seq2seq:copyCopy head branch name to clipboard

Commits

Commits on Oct 24, 2017

• Updated TopKDecoder (IBM#86)

  Show description for 842d8aa
  kylegao91

  authored
  842d8aa

  Copy full SHA for 842d8aa

• Use torchtext from pipy (IBM#87)

  Show description for a32999e
  kylegao91

  authored
  a32999e

  Copy full SHA for a32999e

• attention is not required when only using teacher forcing in decoder (IBM#90)

  Show description for 3f201b8
  

  lifengjin

  authored and

  kylegao91

  committed
  3f201b8

  Copy full SHA for 3f201b8

Commits on Oct 30, 2017

• Updated docs and version.

  kylegao91

  committed
  96af89d

  Copy full SHA for 96af89d

• Fixed code style.

  kylegao91

  committed
  1574e1c

  Copy full SHA for 1574e1c

Commits on Nov 1, 2017

• bugfix (IBM#92)

  Show description for feabc36
  

  Taras Sereda

  authored and

  kylegao91

  committed
  feabc36

  Copy full SHA for feabc36

Commits on Nov 14, 2017

• Removed initial_lr when resuming optimizer with scheduler. (IBM#95)

  kylegao91

  authored
  ed8b90c

  Copy full SHA for ed8b90c

• shuffle the training data (IBM#97)

  Show description for 6ec61bc
  

  KaidongYu

  authored and

  kylegao91

  committed
  6ec61bc

  Copy full SHA for 6ec61bc

Commits on Nov 20, 2017

• fix example of inflate function in TopKDecoer.py (IBM#98)

  Show description for bd3537e
  

  Zhujunnan

  authored and

  kylegao91

  committed
  bd3537e

  Copy full SHA for bd3537e

• Fix hidden_layer size for one-directional decoder (IBM#99)

  Show description for 626842c
  

  dieuwkehupkes

  authored and

  kylegao91

  committed
  626842c

  Copy full SHA for 626842c

• Adapt load to allow CPU loading of GPU models (IBM#100)

  Show description for 97aca03
  

  dieuwkehupkes

  authored and

  kylegao91

  committed
  97aca03

  Copy full SHA for 97aca03

Commits on Dec 4, 2017

• Fix wrong parameter use on DecoderRNN (IBM#103)

  Show description for a3232b0
  

  alejandrodumas

  authored and

  kylegao91

  committed
  a3232b0

  Copy full SHA for a3232b0

• Merge branch 'master' into develop

  Show description for fc584e9
  kylegao91

  committed
  fc584e9

  Copy full SHA for fc584e9

• 0.1.5 + minor fixes (IBM#106)

  Show description for 96c6033
  kylegao91

  authored
  96c6033

  Copy full SHA for 96c6033

Commits on Jan 4, 2018

• Merge branch 'master' into develop

  kylegao91

  committed
  f86e8c7

  Copy full SHA for f86e8c7

• Upgrade to pytorch-0.3.0 (IBM#111)

  Show description for fdbc4a7
  kylegao91

  authored
  fdbc4a7

  Copy full SHA for fdbc4a7

• Make sure tensor contiguous when attention's not used. (IBM#112)

  kylegao91

  authored
  75a8b76

  Copy full SHA for 75a8b76

Commits on Jan 18, 2018

• Implementing the predict_n method. Using the beam search outputs it returns several seqs for a given seq (IBM#116)

  Show description for 38e7e21
  

  cbjuan

  authored and

  kylegao91

  committed
  38e7e21

  Copy full SHA for 38e7e21

Commits on Jan 21, 2018

• Checkpoint after batches not epochs (IBM#119)

  

  tejaswini

  authored and

  kylegao91

  committed
  cbd8d8b

  Copy full SHA for cbd8d8b

Commits on Mar 15, 2018

• add contiguous call to tensor (IBM#127)

  Show description for aef9b9f
  

  alexholdenmiller

  authored and

  kylegao91

  committed
  aef9b9f

  Copy full SHA for aef9b9f

Commits on Apr 2, 2018

• Fixed shape documentation (IBM#131)

  

  kdavis-mozilla

  authored and

  kylegao91

  committed
  4c661ca

  Copy full SHA for 4c661ca

Commits on Apr 25, 2018

• Merge branch 'master' into develop

  kylegao91

  committed
  302c1e8

  Copy full SHA for 302c1e8

• Pytorch 0.4 (IBM#134)

  Show description for 8995987
  kylegao91

  authored
  8995987

  Copy full SHA for 8995987

• Merge remote-tracking branch 'origin/develop' into develop

  kylegao91

  committed
  519db31

  Copy full SHA for 519db31

Commits on May 4, 2018

• Allow using pre-trained embedding (IBM#135)

  kylegao91

  authored
  0e1d875

  Copy full SHA for 0e1d875

• updated docs

  kylegao91

  committed
  ee7d6bd

  Copy full SHA for ee7d6bd

• updated README

  kylegao91

  committed
  4799b7c

  Copy full SHA for 4799b7c

• Release 0.1.6 (IBM#137)

  Show description for fede876
  kylegao91

  authored
  fede876

  Copy full SHA for fede876

• Merge branch 'master' into develop

  kylegao91

  authored
  1d6c6e2

  Copy full SHA for 1d6c6e2

• updated README

  kylegao91

  committed
  17fe235

  Copy full SHA for 17fe235

• Merge branch 'master' of https://github.com/IBM/pytorch-seq2seq

  kylegao91

  committed
  f146087

  Copy full SHA for f146087

Commits on Aug 31, 2018

• minor fixes

  pskrunner14

  committed
  fc570ae

  Copy full SHA for fc570ae

Commits on Sep 1, 2018

• fixed integration test expected cuda bug and minor changes in checkpoint test

  pskrunner14

  committed
  f98c17d

  Copy full SHA for f98c17d

• removes trailing whitespace

  pskrunner14

  committed
  d86e0fa

  Copy full SHA for d86e0fa

• updates README

  pskrunner14

  committed
  447f00a

  Copy full SHA for 447f00a

• removed deprecated warnings and added dropout with 1 layer warning

  pskrunner14

  committed
  554a3b5

  Copy full SHA for 554a3b5

• update README

  pskrunner14

  committed
  4acfa8b

  Copy full SHA for 4acfa8b

• update README

  pskrunner14

  committed
  3ac68ae

  Copy full SHA for 3ac68ae

• changes vagrant file to support pytorch-0.2-linux-python2.7-cpu and updates README

  pskrunner14

  committed
  15a9f30

  Copy full SHA for 15a9f30

• fixed vagrant file

  pskrunner14

  committed
  e3836f8

  Copy full SHA for e3836f8

• updates docs

  pskrunner14

  committed
  f472ca2

  Copy full SHA for f472ca2

• ignore dev folder in docs

  pskrunner14

  committed
  78eab08

  Copy full SHA for 78eab08

• Merge branch 'master' into develop

  pskrunner14

  authored
  e98d50e

  Copy full SHA for e98d50e

• moves docs README to docroot

  pskrunner14

  committed
  8646d6b

  Copy full SHA for 8646d6b

• Merge branch 'develop' of github.com:pskrunner14/pytorch-seq2seq into develop

  pskrunner14

  committed
  a685b43

  Copy full SHA for a685b43

• made ref

  pskrunner14

  committed
  3be3172

  Copy full SHA for 3be3172

• Merge branch 'develop' into copy

  pskrunner14

  authored
  1da2f9c

  Copy full SHA for 1da2f9c

• Update DecoderRNN.py

  Show description for cde0b25
  pskrunner14

  authored
  cde0b25

  Copy full SHA for cde0b25

• Merge pull request #3 from pskrunner14/copy

  Show description for adab17b
  pskrunner14

  authored
  adab17b

  Copy full SHA for adab17b

Commits on Sep 2, 2018

• fixed NoneType has no fields bug

  pskrunner14

  committed
  1a3cd2c

  Copy full SHA for 1a3cd2c

• fixes test loss dataset bug

  pskrunner14

  committed
  f7cb978

  Copy full SHA for f7cb978

• fixes test loss concat list issue

  pskrunner14

  committed
  cb5be01

  Copy full SHA for cb5be01

• fixes sample.py accuracy issue

  pskrunner14

  committed
  04acfb4

  Copy full SHA for 04acfb4

• fixes integration test python3 test fail bug

  pskrunner14

  committed
  ac21962

  Copy full SHA for ac21962

• update docs and README

  pskrunner14

  committed
  eb21a48

  Copy full SHA for eb21a48

• update README

  pskrunner14

  committed
  c2cbd45

  Copy full SHA for c2cbd45

• update docs and README

  pskrunner14

  committed
  7ce607c

  Copy full SHA for 7ce607c

• modified topkdecoder test for more batch sizes

  pskrunner14

  committed
  ca80052

  Copy full SHA for ca80052

• changes topkdecoder test

  pskrunner14

  committed
  a2d9a20

  Copy full SHA for a2d9a20

• changes topkdecoder test

  pskrunner14

  committed
  2322c5b

  Copy full SHA for 2322c5b

• changes topkdecoder test

  pskrunner14

  committed
  51dc6ea

  Copy full SHA for 51dc6ea

• changes topkdecoder test

  pskrunner14

  committed
  20cff5d

  Copy full SHA for 20cff5d

• changes topkdecoder test

  pskrunner14

  committed
  7993fea

  Copy full SHA for 7993fea

• debugging topkdecoder test

  pskrunner14

  committed
  2e82975

  Copy full SHA for 2e82975

• debugging topkdecoder test

  pskrunner14

  committed
  a38254b

  Copy full SHA for a38254b

• debugging topkdecoder test

  pskrunner14

  committed
  993de89

  Copy full SHA for 993de89

• debugging topkdecoder test

  pskrunner14

  committed
  5b54ad1

  Copy full SHA for 5b54ad1

• debugging topkdecoder test

  pskrunner14

  committed
  1df8903

  Copy full SHA for 1df8903

• beam search fails with batch>2

  pskrunner14

  committed
  a148c49

  Copy full SHA for a148c49

• removes deprecated warnings

  pskrunner14

  committed
  7aa9330

  Copy full SHA for 7aa9330

• fixes deprecated warnings

  pskrunner14

  committed
  8229fbe

  Copy full SHA for 8229fbe

• fixes typo

  pskrunner14

  committed
  5f61757

  Copy full SHA for 5f61757

• fixes typo

  pskrunner14

  committed
  539fbc1

  Copy full SHA for 539fbc1

• fixes typo

  pskrunner14

  committed
  59eb3f0

  Copy full SHA for 59eb3f0

Commits on Sep 3, 2018

• refactors modules and makes init changes for copy mech

  pskrunner14

  committed
  c346d33

  Copy full SHA for c346d33

• updated docs and refactored modules

  pskrunner14

  committed
  1d6d251

  Copy full SHA for 1d6d251

Commits on Sep 6, 2018

• branches copy and coverage api to copy

  pskrunner14

  committed
  b733442

  Copy full SHA for b733442

Commits on Sep 21, 2018

• debugging copy decoder dims mismatch

  pskrunner14

  committed
  104e2b9

  Copy full SHA for 104e2b9

• backtracked to prev copy dec def due to dimensional errors

  pskrunner14

  committed
  4fcad39

  Copy full SHA for 4fcad39

Commits on Oct 12, 2018

• fix copy loss

  pskrunner14

  committed
  bf05795

  Copy full SHA for bf05795