mirror of
https://github.com/wassname/TTS.git
synced 2026-09-09 11:16:00 +08:00
pep8 check
This commit is contained in:
Eren Golge
+2
-18
@@ -10,8 +10,8 @@ _mel_basis = None
|
||||
class AudioProcessor(object):
|
||||
|
||||
def __init__(self, sample_rate, num_mels, min_level_db, frame_shift_ms,
|
||||
frame_length_ms, preemphasis, ref_level_db, num_freq, power,
|
||||
griffin_lim_iters=None):
|
||||
frame_length_ms, preemphasis, ref_level_db, num_freq, power,
|
||||
griffin_lim_iters=None):
|
||||
self.sample_rate = sample_rate
|
||||
self.num_mels = num_mels
|
||||
self.min_level_db = min_level_db
|
||||
@@ -23,61 +23,49 @@ class AudioProcessor(object):
|
||||
self.power = power
|
||||
self.griffin_lim_iters = griffin_lim_iters
|
||||
|
||||
|
||||
def save_wav(self, wav, path):
|
||||
wav *= 32767 / max(0.01, np.max(np.abs(wav)))
|
||||
librosa.output.write_wav(path, wav.astype(np.int16), self.sample_rate)
|
||||
|
||||
|
||||
def _linear_to_mel(self, spectrogram):
|
||||
global _mel_basis
|
||||
if _mel_basis is None:
|
||||
_mel_basis = self._build_mel_basis()
|
||||
return np.dot(_mel_basis, spectrogram)
|
||||
|
||||
|
||||
def _build_mel_basis(self, ):
|
||||
n_fft = (self.num_freq - 1) * 2
|
||||
return librosa.filters.mel(self.sample_rate, n_fft, n_mels=self.num_mels)
|
||||
|
||||
|
||||
def _normalize(self, S):
|
||||
return np.clip((S - self.min_level_db) / -self.min_level_db, 0, 1)
|
||||
|
||||
|
||||
def _denormalize(self, S):
|
||||
return (np.clip(S, 0, 1) * -self.min_level_db) + self.min_level_db
|
||||
|
||||
|
||||
def _stft_parameters(self, ):
|
||||
n_fft = (self.num_freq - 1) * 2
|
||||
hop_length = int(self.frame_shift_ms / 1000 * self.sample_rate)
|
||||
win_length = int(self.frame_length_ms / 1000 * self.sample_rate)
|
||||
return n_fft, hop_length, win_length
|
||||
|
||||
|
||||
def _amp_to_db(self, x):
|
||||
return 20 * np.log10(np.maximum(1e-5, x))
|
||||
|
||||
|
||||
def _db_to_amp(self, x):
|
||||
return np.power(10.0, x * 0.05)
|
||||
|
||||
|
||||
def apply_preemphasis(self, x):
|
||||
return signal.lfilter([1, -self.preemphasis], [1], x)
|
||||
|
||||
|
||||
def apply_inv_preemphasis(self, x):
|
||||
return signal.lfilter([1], [1, -self.preemphasis], x)
|
||||
|
||||
|
||||
def spectrogram(self, y):
|
||||
D = self._stft(self.apply_preemphasis(y))
|
||||
S = self._amp_to_db(np.abs(D)) - self.ref_level_db
|
||||
return self._normalize(S)
|
||||
|
||||
|
||||
def inv_spectrogram(self, spectrogram):
|
||||
'''Converts spectrogram to waveform using librosa'''
|
||||
S = self._denormalize(spectrogram)
|
||||
@@ -85,7 +73,6 @@ class AudioProcessor(object):
|
||||
# Reconstruct phase
|
||||
return self.apply_inv_preemphasis(self._griffin_lim(S ** self.power))
|
||||
|
||||
|
||||
def _griffin_lim(self, S):
|
||||
'''librosa implementation of Griffin-Lim
|
||||
Based on https://github.com/librosa/librosa/issues/434
|
||||
@@ -98,13 +85,11 @@ class AudioProcessor(object):
|
||||
y = self._istft(S_complex * angles)
|
||||
return y
|
||||
|
||||
|
||||
def melspectrogram(self, y):
|
||||
D = self._stft(self.apply_preemphasis(y))
|
||||
S = self._amp_to_db(self._linear_to_mel(np.abs(D))) - self.ref_level_db
|
||||
return self._normalize(S)
|
||||
|
||||
|
||||
def _stft(self, y):
|
||||
n_fft, hop_length, win_length = self._stft_parameters()
|
||||
return librosa.stft(y=y, n_fft=n_fft, hop_length=hop_length, win_length=win_length)
|
||||
@@ -113,7 +98,6 @@ class AudioProcessor(object):
|
||||
_, hop_length, win_length = self._stft_parameters()
|
||||
return librosa.istft(y, hop_length=hop_length, win_length=win_length)
|
||||
|
||||
|
||||
def find_endpoint(self, wav, threshold_db=-40, min_silence_sec=0.8):
|
||||
window_length = int(self.sample_rate * min_silence_sec)
|
||||
hop_length = int(window_length / 4)
|
||||
|
||||
+4
-2
@@ -17,11 +17,13 @@ def prepare_data(inputs):
|
||||
def _pad_tensor(x, length):
|
||||
_pad = 0
|
||||
assert x.ndim == 2
|
||||
x = np.pad(x, [[0, 0], [0, length - x.shape[1]]], mode='constant', constant_values=_pad)
|
||||
x = np.pad(x, [[0, 0], [0, length - x.shape[1]]],
|
||||
mode='constant', constant_values=_pad)
|
||||
return x
|
||||
|
||||
|
||||
def prepare_tensor(inputs, out_steps):
|
||||
max_len = max((x.shape[1] for x in inputs)) + 1 # zero-frame
|
||||
max_len = max((x.shape[1] for x in inputs)) + 1 # zero-frame
|
||||
remainder = max_len % out_steps
|
||||
pad_len = max_len + (out_steps - remainder) if remainder > 0 else max_len
|
||||
return np.stack([_pad_tensor(x, pad_len) for x in inputs])
|
||||
|
||||
@@ -19,7 +19,7 @@ class AttrDict(dict):
|
||||
def load_config(config_path):
|
||||
config = AttrDict()
|
||||
config.update(json.load(open(config_path, "r")))
|
||||
return config
|
||||
return config
|
||||
|
||||
|
||||
def create_experiment_folder(root_path):
|
||||
@@ -56,7 +56,7 @@ def _trim_model_state_dict(state_dict):
|
||||
|
||||
new_state_dict = OrderedDict()
|
||||
for k, v in state_dict.items():
|
||||
name = k[7:] # remove `module.`
|
||||
name = k[7:] # remove `module.`
|
||||
new_state_dict[name] = v
|
||||
return new_state_dict
|
||||
|
||||
@@ -90,7 +90,8 @@ def save_best_model(model, optimizer, model_loss, best_loss, out_path,
|
||||
best_loss = model_loss
|
||||
bestmodel_path = 'best_model.pth.tar'
|
||||
bestmodel_path = os.path.join(out_path, bestmodel_path)
|
||||
print("\n | > Best model saving with loss {0:.2f} : {1:}".format(model_loss, bestmodel_path))
|
||||
print("\n | > Best model saving with loss {0:.2f} : {1:}".format(
|
||||
model_loss, bestmodel_path))
|
||||
torch.save(state, bestmodel_path)
|
||||
return best_loss
|
||||
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
#-*- coding: utf-8 -*-
|
||||
# -*- coding: utf-8 -*-
|
||||
|
||||
import re
|
||||
from TTS.utils.text import cleaners
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
#-*- coding: utf-8 -*-
|
||||
# -*- coding: utf-8 -*-
|
||||
|
||||
|
||||
'''
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
#-*- coding: utf-8 -*-
|
||||
# -*- coding: utf-8 -*-
|
||||
|
||||
|
||||
import re
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
#-*- coding: utf-8 -*-
|
||||
# -*- coding: utf-8 -*-
|
||||
|
||||
import inflect
|
||||
import re
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
#-*- coding: utf-8 -*-
|
||||
# -*- coding: utf-8 -*-
|
||||
|
||||
|
||||
'''
|
||||
|
||||
+1
-1
@@ -5,7 +5,7 @@ import matplotlib.pyplot as plt
|
||||
|
||||
|
||||
def plot_alignment(alignment, info=None):
|
||||
fig, ax = plt.subplots(figsize=(16,10))
|
||||
fig, ax = plt.subplots(figsize=(16, 10))
|
||||
im = ax.imshow(alignment.T, aspect='auto', origin='lower',
|
||||
interpolation='none')
|
||||
fig.colorbar(im, ax=ax)
|
||||
|
||||
Reference in New Issue
Block a user