File size: 620 Bytes
0b32ad6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 |
from .base import SequentialDataPipe
from .common_pipes import LoadAudio, Phonemize, SetOutputKeys
class Speech2PhonemePipe(SequentialDataPipe):
"""
each item in the input dataset should have:
wav_path: str
transcription: str
"""
def __init__(self):
output_keys = dict(
x="wav",
x_len="wav_len",
labels="phonemized_text",
class_ids="tokenized_text",
unique_name="id",
)
super().__init__(
LoadAudio(),
Phonemize(),
SetOutputKeys(output_keys=output_keys),
)
|