Skip to content

OpenNMT-tf 2.1.0

Compare
Choose a tag to compare
@guillaumekln guillaumekln released this 10 Oct 12:38
· 522 commits to master since this release

OpenNMT-tf 2.1.0

New features

  • onmt-build-vocab script can now train a SentencePiece model and vocabulary from raw data
  • Enable automatic model export during evaluation with export_on_best parameter
  • Add perplexity in evaluation metrics
  • Extend tokenization configuration to support in-graph tokenizers (currently SpaceTokenizer and CharacterTokenizer)
  • Parameter decoder_subword_token_is_spacer to configure the type of decoder_subword_token
  • [API] Support tf.RaggedTensor in tokenizer API

Fixes and improvements

  • Fix early stopping logic
  • Support spacer decoder_subword_token that is used as a suffix
  • Improve errors when --model or --model_type options are invalid