auracaster/piper
3
0
Fork 0
mirror of https://github.com/pstrueb/piper.git synced 2026-04-18 06:15:30 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add --speaker-id to preprocess to force a speaker id

Browse Source
This commit is contained in:
Michael Hansen
2023-03-17 12:03:37 -05:00
parent 8873638771
commit 44c2cf1ee9
1 changed files with 20 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
src/python/larynx_train/preprocess.py
View File

@@ -45,11 +45,18 @@ def main() -> None:
parser.add_argument(
"--single-speaker", action="store_true", help="Force single speaker dataset"
)
parser.add_argument(
"--speaker-id", type=int, help="Add speaker id to single speaker dataset"
)
parser.add_argument(
"--debug", action="store_true", help="Print DEBUG messages to the console"
)
args = parser.parse_args()
if args.single_speaker and (args.speaker_id is not None):
_LOGGER.fatal("--single-speaker and --speaker-id cannot both be provided")
return
level = logging.DEBUG if args.debug else logging.INFO
logging.basicConfig(level=level)
logging.getLogger().setLevel(level)
@@ -78,7 +85,7 @@ def main() -> None:
_LOGGER.debug("Counting number of speakers/utterances in the dataset")
speaker_counts: Counter[str] = Counter()
num_utterances = 0
for utt in make_dataset(args.input_dir, args.single_speaker):
for utt in make_dataset(args.input_dir, args.single_speaker, args.speaker_id):
speaker = utt.speaker or ""
speaker_counts[speaker] += 1
num_utterances += 1
@@ -146,7 +153,8 @@ def main() -> None:
)
with open(args.output_dir / "dataset.jsonl", "w", encoding="utf-8") as dataset_file:
for utt_batch in batched(
make_dataset(args.input_dir, args.single_speaker), batch_size
make_dataset(args.input_dir, args.single_speaker, args.speaker_id),
batch_size,
):
queue_in.put(utt_batch)
@@ -231,7 +239,9 @@ class PathEncoder(json.JSONEncoder):
return super().default(o)
def ljspeech_dataset(dataset_dir: Path, is_single_speaker: bool) -> Iterable[Utterance]:
def ljspeech_dataset(
dataset_dir: Path, is_single_speaker: bool, speaker_id: Optional[int] = None
) -> Iterable[Utterance]:
# filename|speaker|text
# speaker is optional
metadata_path = dataset_dir / "metadata.csv"
@@ -271,15 +281,19 @@ def ljspeech_dataset(dataset_dir: Path, is_single_speaker: bool) -> Iterable[Utt
_LOGGER.warning("Missing %s", filename)
continue
yield Utterance(text=text, audio_path=wav_path, speaker=speaker)
yield Utterance(
text=text, audio_path=wav_path, speaker=speaker, speaker_id=speaker_id
)
def mycroft_dataset(dataset_dir: Path, is_single_speaker: bool) -> Iterable[Utterance]:
def mycroft_dataset(
dataset_dir: Path, is_single_speaker: bool, speaker_id: Optional[int] = None
) -> Iterable[Utterance]:
for info_path in dataset_dir.glob("*.info"):
wav_path = info_path.with_suffix(".wav")
if wav_path.exists():
text = info_path.read_text(encoding="utf-8").strip()
yield Utterance(text=text, audio_path=wav_path)
yield Utterance(text=text, audio_path=wav_path, speaker_id=speaker_id)
# -----------------------------------------------------------------------------