Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

linter fix

This commit is contained in:
Eren Golge 2019-09-11 10:39:59 +02:00
parent 609d8efa69
commit d45d963dc1
3 changed files with 23 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
train.py
View File

@ -305,9 +305,9 @@ def evaluate(model, criterion, criterion_st, ap, global_step, epoch):
speaker_mapping = load_speaker_mapping(OUT_PATH) speaker_mapping = load_speaker_mapping(OUT_PATH)
model.eval() model.eval()
epoch_time = 0 epoch_time = 0
eval_values_dict = {'avg_postnet_loss' : 0, eval_values_dict = {'avg_postnet_loss': 0,
'avg_decoder_loss' : 0, 'avg_decoder_loss': 0,
'avg_stop_loss' : 0, 'avg_stop_loss': 0,
'avg_align_score': 0} 'avg_align_score': 0}
keep_avg = KeepAverage() keep_avg = KeepAverage()
keep_avg.add_values(eval_values_dict) keep_avg.add_values(eval_values_dict)
@ -401,18 +401,19 @@ def evaluate(model, criterion, criterion_st, ap, global_step, epoch):
if c.stopnet: if c.stopnet:
stop_loss = reduce_tensor(stop_loss.data, num_gpus) stop_loss = reduce_tensor(stop_loss.data, num_gpus)
keep_avg.update_values({'avg_postnet_loss' : float(postnet_loss.item()), keep_avg.update_values({'avg_postnet_loss': float(postnet_loss.item()),
'avg_decoder_loss' : float(decoder_loss.item()), 'avg_decoder_loss': float(decoder_loss.item()),
'avg_stop_loss' : float(stop_loss.item())}) 'avg_stop_loss': float(stop_loss.item())})
if num_iter % c.print_step == 0: if num_iter % c.print_step == 0:
print( print(
" | > TotalLoss: {:.5f} PostnetLoss: {:.5f} - {:.5f} DecoderLoss:{:.5f} - {:.5f} " " | > TotalLoss: {:.5f} PostnetLoss: {:.5f} - {:.5f} DecoderLoss:{:.5f} - {:.5f} "
"StopLoss: {:.5f} - {:.5f} AlignScore: {:.4f} : {:.4f}".format(loss.item(), "StopLoss: {:.5f} - {:.5f} AlignScore: {:.4f} : {:.4f}".format(
postnet_loss.item(), keep_avg['avg_postnet_loss'], loss.item(),
decoder_loss.item(), keep_avg['avg_decoder_loss'], postnet_loss.item(), keep_avg['avg_postnet_loss'],
stop_loss.item(), keep_avg['avg_stop_loss'], decoder_loss.item(), keep_avg['avg_decoder_loss'],
align_score.item(), keep_avg['avg_align_score']), stop_loss.item(), keep_avg['avg_stop_loss'],
align_score.item(), keep_avg['avg_align_score']),
flush=True) flush=True)
if args.rank == 0: if args.rank == 0:

7
utils/generic_utils.py
View File

@ -31,7 +31,8 @@ def load_config(config_path):
def get_git_branch(): def get_git_branch():
try: try:
out = subprocess.check_output(["git", "branch"]).decode("utf8") out = subprocess.check_output(["git", "branch"]).decode("utf8")
current = next(line for line in out.split("\n") if line.startswith("*")) current = next(line for line in out.split(
"\n") if line.startswith("*"))
current.replace("* ", "") current.replace("* ", "")
except subprocess.CalledProcessError: except subprocess.CalledProcessError:
current = "inside_docker" current = "inside_docker"
@ -333,7 +334,8 @@ class KeepAverage():
self.avg_values[name] = 0.99 * self.avg_values[name] + 0.01 * value self.avg_values[name] = 0.99 * self.avg_values[name] + 0.01 * value
self.iters[name] += 1 self.iters[name] += 1
else: else:
self.avg_values[name] = self.avg_values[name] * self.iters[name] + value self.avg_values[name] = self.avg_values[name] * \
self.iters[name] + value
self.iters[name] += 1 self.iters[name] += 1
self.avg_values[name] /= self.iters[name] self.avg_values[name] /= self.iters[name]
@ -344,4 +346,3 @@ class KeepAverage():
def update_values(self, value_dict): def update_values(self, value_dict):
for key, value in value_dict.items(): for key, value in value_dict.items():
self.update_value(key, value) self.update_value(key, value)

8
utils/measures.py
View File

@ -1,6 +1,3 @@
import torch
import numpy as np
def alignment_diagonal_score(alignments): def alignment_diagonal_score(alignments):
""" """
@ -12,8 +9,3 @@ def alignment_diagonal_score(alignments):
alignments : batch x decoder_steps x encoder_steps alignments : batch x decoder_steps x encoder_steps
""" """
return alignments.max(dim=1)[0].mean(dim=1).mean(dim=0) return alignments.max(dim=1)[0].mean(dim=1).mean(dim=0)