extract_lsr2_vocab.py

from TranslatorTrainer import *

vocab_filepaths = (
    'data/lsr2_phoneme_vocab.pth',
    'data/lsr2_text_char_vocab.pth'
)

trainer = TranslatorTrainer(
    word_tokenize=False,
    input_char_map=options.char_map,
    output_char_map=options.text_char_map,
    dataset_type=Datasets.LRS2
)

# trainer.train()
trainer.save_vocabs(*vocab_filepaths)