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

Merge branch 'larger-attn-larger-model'

This commit is contained in:
Eren 2018-09-17 13:58:57 +02:00
parent 48035a329f 4b08c3a13f
commit 41c06029b8
2 changed files with 10 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
layers/attention.py
View File

@ -37,8 +37,8 @@ class LocationSensitiveAttention(nn.Module):
annot_dim, annot_dim,
query_dim, query_dim,
attn_dim, attn_dim,
kernel_size=7, kernel_size=31,
filters=20): filters=32):
super(LocationSensitiveAttention, self).__init__() super(LocationSensitiveAttention, self).__init__()
self.kernel_size = kernel_size self.kernel_size = kernel_size
self.filters = filters self.filters = filters

9
train.py
View File

@ -315,10 +315,15 @@ def evaluate(model, criterion, criterion_st, data_loader, ap, current_step):
try: try:
wav, linear_spec, alignments = synthesis(model, ap, test_sentence, wav, linear_spec, alignments = synthesis(model, ap, test_sentence,
use_cuda, c.text_cleaner) use_cuda, c.text_cleaner)
file_path = os.path.join(AUDIO_PATH, str(current_step))
os.makedirs(file_path, exist_ok=True)
file_path = os.path.join(file_path, "TestSentence_{}.wav".format(idx))
ap.save_wav(wav, file_path)
wav_name = 'TestSentences/{}'.format(idx) wav_name = 'TestSentences/{}'.format(idx)
tb.add_audio( tb.add_audio(
wav_name, wav, current_step, sample_rate=c.sample_rate) wav_name, wav, current_step, sample_rate=c.sample_rate)
align_img = alignments[0].data.cpu().numpy() align_img = alignments[0].data.cpu().numpy()
linear_spec = plot_spectrogram(linear_spec, ap) linear_spec = plot_spectrogram(linear_spec, ap)
align_img = plot_alignment(align_img) align_img = plot_alignment(align_img)
@ -466,6 +471,8 @@ if __name__ == '__main__':
OUT_PATH = os.path.join(_, c.output_path) OUT_PATH = os.path.join(_, c.output_path)
OUT_PATH = create_experiment_folder(OUT_PATH, c.model_name, args.debug) OUT_PATH = create_experiment_folder(OUT_PATH, c.model_name, args.debug)
CHECKPOINT_PATH = os.path.join(OUT_PATH, 'checkpoints') CHECKPOINT_PATH = os.path.join(OUT_PATH, 'checkpoints')
AUDIO_PATH = os.path.join(OUT_PATH, 'test_audios')
os.mkdir(AUDIO_PATH)
shutil.copyfile(args.config_path, os.path.join(OUT_PATH, 'config.json')) shutil.copyfile(args.config_path, os.path.join(OUT_PATH, 'config.json'))
# setup tensorboard # setup tensorboard