Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Jul 12, 2017

• Update public code links.

  Show description for 70126d8
  oahziur

  committed
  70126d8

  Copy full SHA for 70126d8

• Update Links in Readme.

  Show description for 7bd686b
  oahziur

  committed
  7bd686b

  Copy full SHA for 7bd686b

• Remove extract vocab from corpus, --eval_only and rename path_to_standard_hparams to hparams_path.

  Show description for fb9aeb4
  oahziur

  committed
  fb9aeb4

  Copy full SHA for fb9aeb4

• Add installation instructions and update syntax highlighting.

  Show description for d2e4e8b
  

  ebrevdo

  authored and

  oahziur

  committed
  d2e4e8b

  Copy full SHA for d2e4e8b

• Add English-German results, compare with other systems. Update acknowledgement.

  Show description for a837a34
  

  lmthang

  authored and

  oahziur

  committed
  a837a34

  Copy full SHA for a837a34

• Update standard_hparams to match the experiment results

  Show description for e6dd7bd
  oahziur

  committed
  e6dd7bd

  Copy full SHA for e6dd7bd

• Update experimental results in README.md

  Show description for 19849b0
  

  lmthang

  authored and

  oahziur

  committed
  19849b0

  Copy full SHA for 19849b0

• Update minimum tf version to 1.2.1.

  Show description for d91185f
  oahziur

  committed
  d91185f

  Copy full SHA for d91185f

• Add num_train_steps to the updated_keys in ensure_compatible_hparams()

  Show description for 329dd49
  

  lmthang

  authored and

  oahziur

  committed
  329dd49

  Copy full SHA for 329dd49

Commits on Jul 10, 2017

• import nest from tensorflow.python.util for tf 1.2 compatibility.

  Show description for 07d3d78
  oahziur

  committed
  07d3d78

  Copy full SHA for 07d3d78

• Increase buffer size for iterator

  Show description for 33d910d
  

  lmthang

  authored and

  oahziur

  committed
  33d910d

  Copy full SHA for 33d910d

• Update tf.initialize_all_tables() to tf.table_initializer().

  Show description for a7dfc5a
  oahziur

  committed
  a7dfc5a

  Copy full SHA for a7dfc5a

• Update the python invoke command for running NMT locally.

  Show description for 97414aa
  oahziur

  committed
  97414aa

  Copy full SHA for 97414aa

• Add an optional target_session to train.train() and some clean-ups.

  Show description for 7ee7bed
  

  lmthang

  authored and

  oahziur

  committed
  7ee7bed

  Copy full SHA for 7ee7bed

• Add length_penalty_weight and fix beam search in _sample_decode.

  Show description for 77b20b3
  oahziur

  committed
  77b20b3

  Copy full SHA for 77b20b3

• Add eval_best_only mode to eval saved best model.

  Show description for 7a6a784
  oahziur

  committed
  7a6a784

  Copy full SHA for 7a6a784

• Fix infer evaluation.

  Show description for ca81f21
  oahziur

  committed
  ca81f21

  Copy full SHA for ca81f21

• Add gnmt_v2 attention architecture.

  Show description for c053893
  oahziur

  committed
  c053893

  Copy full SHA for c053893

• Add pass_hidden_state as an option & bug fix on loading hparams.

  Show description for 3135d41
  oahziur

  committed
  3135d41

  Copy full SHA for 3135d41

• Check if unk, sos, eos are in a vocab file. If not, create.

  Show description for 842c435
  

  lmthang

  authored and

  oahziur

  committed
  842c435

  Copy full SHA for 842c435

• Remove unused hparams and clean up.

  Show description for dc947f0
  oahziur

  committed
  dc947f0

  Copy full SHA for dc947f0

• Save hparams for submodel directories, remove best_ppl / best_ppl_dir from

  Show description for 345e2fc
  

  lmthang

  authored and

  oahziur

  committed
  345e2fc

  Copy full SHA for 345e2fc

• Skip training data elements after job restarts.

  Show description for 498b3fe
  oahziur

  committed
  498b3fe

  Copy full SHA for 498b3fe

• Add inference flags (inference_output_file, inference_ref_file), rename

  Show description for b2d88c3
  

  lmthang

  authored and

  oahziur

  committed
  b2d88c3

  Copy full SHA for b2d88c3

• Remove outer training loop and add sample_decode in internal_eval().

  Show description for 1bfe14c
  oahziur

  committed
  1bfe14c

  Copy full SHA for 1bfe14c

• Remove unused functions and ignore_map.

  Show description for 1fccf25
  oahziur

  committed
  1fccf25

  Copy full SHA for 1fccf25

• Copy downloading scripts from seq2seq tutorial.

  Show description for 1ba8f93
  oahziur

  committed
  1ba8f93

  Copy full SHA for 1ba8f93

• Fixes the invalid markdown references in NMT tutorial.

  Show description for 534d25f
  oahziur

  committed
  534d25f

  Copy full SHA for 534d25f

• Move seq2seq tutorial to README.md

  Show description for 04d4d5f
  

  ebrevdo

  authored and

  oahziur

  committed
  04d4d5f

  Copy full SHA for 04d4d5f

• Delete old README.md

  Show description for e4208fe
  

  ebrevdo

  authored and

  oahziur

  committed
  e4208fe

  Copy full SHA for e4208fe

• Add tutorial .md file.

  Show description for 014166f
  oahziur

  committed
  014166f

  Copy full SHA for 014166f

• Remove gradient clip pattern, gradient clip value, task, and pretrained embedding related FLAGS.

  Show description for d7d2337
  oahziur

  committed
  d7d2337

  Copy full SHA for d7d2337

• Remove number of epoch in training loop.

  Show description for c691c82
  oahziur

  committed
  c691c82

  Copy full SHA for c691c82

• Add --path_to_standard_hparams flag.

  Show description for 9ef191a
  oahziur

  committed
  9ef191a

  Copy full SHA for 9ef191a

• Remove source_reverse, classification and compatible FLAG.

  Show description for 058828c
  oahziur

  committed
  058828c

  Copy full SHA for 058828c

Pagination

Previous

Next