kxngluca/DL-Art-School
1
0
Fork 0
forked from mrq/DL-Art-School
Code Pull Requests Activity

nv_tacotron_dataset: allow it to load conditioning signals

Browse Source
This commit is contained in:
James Betker 2021-12-02 22:14:44 -07:00
parent 07b0124712
commit 702607556d
1 changed files with 56 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

72
codes/data/audio/nv_tacotron_dataset.py
View File

@ -11,6 +11,7 @@ from tqdm import tqdm
import models.tacotron2.layers as layers import models.tacotron2.layers as layers
from data.audio.unsupervised_audio_dataset import load_audio from data.audio.unsupervised_audio_dataset import load_audio
from data.util import find_files_of_type, is_audio_file
from models.tacotron2.taco_utils import load_wav_to_torch, load_filepaths_and_text from models.tacotron2.taco_utils import load_wav_to_torch, load_filepaths_and_text
from models.tacotron2.text import text_to_sequence from models.tacotron2.text import text_to_sequence
@ -50,13 +51,18 @@ class TextWavLoader(torch.utils.data.Dataset):
fetcher_mode = [fetcher_mode] fetcher_mode = [fetcher_mode]
assert len(self.path) == len(fetcher_mode) assert len(self.path) == len(fetcher_mode)
self.load_conditioning = opt_get(hparams, ['load_conditioning'], False)
self.conditioning_candidates = opt_get(hparams, ['num_conditioning_candidates'], 3)
self.conditioning_length = opt_get(hparams, ['conditioning_length'], 44100)
self.audiopaths_and_text = [] self.audiopaths_and_text = []
for p, fm in zip(self.path, fetcher_mode): for p, fm in zip(self.path, fetcher_mode):
if fm == 'lj' or fm == 'libritts': if fm == 'lj' or fm == 'libritts':
fetcher_fn = load_filepaths_and_text fetcher_fn = load_filepaths_and_text
elif fm == 'mozilla_cv': elif fm == 'mozilla_cv':
assert not self.load_conditioning # Conditioning inputs are incompatible with mozilla_cv
fetcher_fn = load_mozilla_cv fetcher_fn = load_mozilla_cv
elif fm == 'voxpopuli': elif fm == 'voxpopuli':
assert not self.load_conditioning # Conditioning inputs are incompatible with voxpopuli
fetcher_fn = load_voxpopuli fetcher_fn = load_voxpopuli
else: else:
raise NotImplementedError() raise NotImplementedError()
@ -83,12 +89,32 @@ class TextWavLoader(torch.utils.data.Dataset):
text_norm = torch.IntTensor(text_to_sequence(text, self.text_cleaners)) text_norm = torch.IntTensor(text_to_sequence(text, self.text_cleaners))
return text_norm return text_norm
def load_conditioning_candidates(self, path):
candidates = find_files_of_type('img', os.path.dirname(path), qualifier=is_audio_file)[0]
assert len(candidates) < 50000 # Sanity check to ensure we aren't loading "related files" that aren't actually related.
if len(candidates) == 0:
print(f"No conditioning candidates found for {path} (not even the clip itself??)")
raise NotImplementedError()
# Sample with replacement. This can get repeats, but more conveniently handles situations where there are not enough candidates.
related_clips = []
for k in range(self.conditioning_candidates):
rel_clip = load_audio(random.choice(candidates), self.sample_rate)
gap = rel_clip.shape[-1] - self.conditioning_length
if gap < 0:
rel_clip = F.pad(rel_clip, pad=(0, abs(gap)))
elif gap > 0:
rand_start = random.randint(0, gap)
rel_clip = rel_clip[:, rand_start:rand_start+self.conditioning_length]
related_clips.append(rel_clip)
return torch.stack(related_clips, dim=0)
def __getitem__(self, index): def __getitem__(self, index):
try: #try:
tseq, wav, text, path = self.get_wav_text_pair(self.audiopaths_and_text[index]) tseq, wav, text, path = self.get_wav_text_pair(self.audiopaths_and_text[index])
except: cond = self.load_conditioning_candidates(self.audiopaths_and_text[index][0]) if self.load_conditioning else None
print(f"error loadding {self.audiopaths_and_text[index][0]}") #except:
return self[index+1] # print(f"error loading {self.audiopaths_and_text[index][0]}")
# return self[index+1]
if wav is None or \ if wav is None or \
(self.max_wav_len is not None and wav.shape[-1] > self.max_wav_len) or \ (self.max_wav_len is not None and wav.shape[-1] > self.max_wav_len) or \
(self.max_text_len is not None and tseq.shape[0] > self.max_text_len): (self.max_text_len is not None and tseq.shape[0] > self.max_text_len):
@ -105,7 +131,7 @@ class TextWavLoader(torch.utils.data.Dataset):
wav = F.pad(wav, (0, self.max_wav_len - wav.shape[-1])) wav = F.pad(wav, (0, self.max_wav_len - wav.shape[-1]))
if tseq.shape[0] != self.max_text_len: if tseq.shape[0] != self.max_text_len:
tseq = F.pad(tseq, (0, self.max_text_len - tseq.shape[0])) tseq = F.pad(tseq, (0, self.max_text_len - tseq.shape[0]))
return { res = {
'real_text': text, 'real_text': text,
'padded_text': tseq, 'padded_text': tseq,
'text_lengths': torch.tensor(orig_text_len, dtype=torch.long), 'text_lengths': torch.tensor(orig_text_len, dtype=torch.long),
@ -113,7 +139,10 @@ class TextWavLoader(torch.utils.data.Dataset):
'wav_lengths': torch.tensor(orig_output, dtype=torch.long), 'wav_lengths': torch.tensor(orig_output, dtype=torch.long),
'filenames': path 'filenames': path
} }
return tseq, wav, path, text if self.load_conditioning:
res['conditioning'] = cond
return res
return tseq, wav, path, text, cond
def __len__(self): def __len__(self):
return len(self.audiopaths_and_text) return len(self.audiopaths_and_text)
@ -138,11 +167,15 @@ class TextMelCollate():
text_padded.zero_() text_padded.zero_()
filenames = [] filenames = []
real_text = [] real_text = []
conds = []
for i in range(len(ids_sorted_decreasing)): for i in range(len(ids_sorted_decreasing)):
text = batch[ids_sorted_decreasing[i]][0] text = batch[ids_sorted_decreasing[i]][0]
text_padded[i, :text.size(0)] = text text_padded[i, :text.size(0)] = text
filenames.append(batch[ids_sorted_decreasing[i]][2]) filenames.append(batch[ids_sorted_decreasing[i]][2])
real_text.append(batch[ids_sorted_decreasing[i]][3]) real_text.append(batch[ids_sorted_decreasing[i]][3])
c = batch[ids_sorted_decreasing[i]][4]
if c is not None:
conds.append(c)
# Right zero-pad wav # Right zero-pad wav
num_wavs = batch[0][1].size(0) num_wavs = batch[0][1].size(0)
@ -157,7 +190,7 @@ class TextMelCollate():
wav_padded[i, :, :wav.size(1)] = wav wav_padded[i, :, :wav.size(1)] = wav
output_lengths[i] = wav.size(1) output_lengths[i] = wav.size(1)
return { res = {
'padded_text': text_padded, 'padded_text': text_padded,
'text_lengths': input_lengths, 'text_lengths': input_lengths,
'wav': wav_padded, 'wav': wav_padded,
@ -165,21 +198,25 @@ class TextMelCollate():
'filenames': filenames, 'filenames': filenames,
'real_text': real_text, 'real_text': real_text,
} }
if len(conds) > 0:
res['conditioning'] = torch.stack(conds)
return res
if __name__ == '__main__': if __name__ == '__main__':
batch_sz = 32 batch_sz = 8
params = { params = {
'mode': 'nv_tacotron', 'mode': 'nv_tacotron',
'path': ['Z:\\bigasr_dataset\\libritts\\test-clean_list.txt'], 'path': ['Z:\\bigasr_dataset\\libritts\\test-clean_list.txt'],
'phase': 'train', 'phase': 'train',
'n_workers': 1, 'n_workers': 0,
'batch_size': batch_sz, 'batch_size': batch_sz,
'fetcher_mode': ['libritts'], 'fetcher_mode': ['libritts'],
'needs_collate': True, 'needs_collate': True,
'max_wav_length': 256000, 'max_wav_length': 256000,
'max_text_length': 200, 'max_text_length': 200,
'sample_rate': 22050, 'sample_rate': 22050,
'load_conditioning': True,
} }
from data import create_dataset, create_dataloader from data import create_dataset, create_dataloader
@ -187,9 +224,12 @@ if __name__ == '__main__':
dl = create_dataloader(ds, params, collate_fn=c) dl = create_dataloader(ds, params, collate_fn=c)
i = 0 i = 0
m = None m = None
for k in range(1000): for i, b in tqdm(enumerate(dl)):
for i, b in tqdm(enumerate(dl)): if i > 5:
w = b['wav'] break
for ib in range(batch_sz): w = b['wav']
print(f'{i} {ib} {b["real_text"][ib]}') for ib in range(batch_sz):
torchaudio.save(f'{i}_clip_{ib}.wav', b['wav'][ib], ds.sample_rate) print(f'{i} {ib} {b["real_text"][ib]}')
torchaudio.save(f'{i}_clip_{ib}.wav', b['wav'][ib], ds.sample_rate)
for c in range(3):
torchaudio.save(f'{i}_clip_{ib}_cond{c}.wav', b['conditioning'][ib, c], ds.sample_rate)