nnmnkwii.io.hts.load_question_set(qs_file_name, append_hat_for_LL=True, convert_svs_pattern=True)[source]

Load HTS-style question and convert it to binary/continuous feature extraction regexes.

This code was taken from Merlin.

  • qs_file_name (str) – Input HTS-style question file path

  • append_hat_for_LL (bool) – Append ^ for LL regex search. Note that the most left context is assumed to be phoneme identity before the previous phoneme (i.e. LL-xx). This parameter should be False for the HTS-demo_NIT-SONG070-F001 demo.

  • convert_svs_pattern (bool) – Convert SVS specific patterns.


Binary/numeric feature extraction regexes.

Return type

(binary_dict, numeric_dict)


>>> from nnmnkwii.io import hts
>>> from nnmnkwii.util import example_question_file
>>> binary_dict, numeric_dict = hts.load_question_set(example_question_file())