v0.3.7

• Wrapped ONNX Inference Session to allow pickling, useful for model serialization with tokenizers and/or parallel processes where serializability is needed.

v0.3.5

• Fixes #16.
• Pinned NLTK version due to backward incompatibility with >=3.8.1

v0.3.4

Code refactors as suggested by Pylint.

v0.3.3

Fix ONNX Runtime-related issues.

v0.3.2

• Swapped word_tokenize to TweetTokenizer.

v0.3.1

Fixed minor rule-based g2p ordering issue.

v0.3.0

Minor improvements to rule-based g2p conversion:

• Handled plurals e.g. (anak-anak)
• Fixed g2p conversion ordering between x (kh grapheme) and ʔs (x grapheme)

v0.0.2

• Split each phoneme as its own token in a list.
• Followed the IPA convention for phonemes, in particular, we followed the list provided in Wikipedia.

v0.0.1

removed numpy from dependency