def build_from_path(in_dir,
out_dir,
silence_threshold,
fft_size,
num_workers=1,
tqdm=lambda x: x):
executor = ProcessPoolExecutor(max_workers=num_workers)
futures = []
speakers = cmu_arctic.available_speakers
wd = cmu_arctic.WavFileDataSource(in_dir, speakers=speakers)
wav_paths = wd.collect_files()
speaker_ids = wd.labels
for index, (speaker_id, wav_path) in enumerate(zip(speaker_ids,
wav_paths)):
# In case of test file
if not os.path.exists(wav_path):
continue
futures.append(
executor.submit(
partial(_process_utterance, out_dir, index + 1, speaker_id,
wav_path, "N/A", silence_threshold, fft_size)))
return [future.result() for future in tqdm(futures)]
def test_cmu_arctic_dummy():
data_source = cmu_arctic.WavFileDataSource("dummy", speakers=["clb"])
@raises(ValueError)
def __test_invalid_speaker():
data_source = cmu_arctic.WavFileDataSource("dummy", speakers=["test"])
@raises(RuntimeError)
def __test_nodir(data_source):
data_source.collect_files()
__test_invalid_speaker()
__test_nodir(data_source)
def build_from_path(in_dir, out_dir, num_workers=1, tqdm=lambda x: x):
executor = ProcessPoolExecutor(max_workers=num_workers)
futures = []
speakers = cmu_arctic.available_speakers
wd = cmu_arctic.WavFileDataSource(in_dir, speakers=speakers)
wav_paths = wd.collect_files()
speaker_ids = wd.labels
for index, (speaker_id, wav_path) in enumerate(
zip(speaker_ids, wav_paths)):
futures.append(executor.submit(
partial(_process_utterance, out_dir, index + 1, speaker_id, wav_path, "N/A")))
return [future.result() for future in tqdm(futures)]
def __test_invalid_speaker():
data_source = cmu_arctic.WavFileDataSource("dummy", speakers=["test"])