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

removed shuffling of data in the preprocessor, uniform indentation

This commit is contained in:
Thomas Werkmeister 2019-04-25 11:46:11 +02:00
parent 41e3e42989
commit f279fe9e8b
1 changed files with 46 additions and 51 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

93
datasets/preprocess.py
View File

@ -1,5 +1,4 @@
import os import os
import random
def tts_cache(root_path, meta_file): def tts_cache(root_path, meta_file):
@ -9,8 +8,8 @@ def tts_cache(root_path, meta_file):
with open(txt_file, 'r', encoding='utf8') as f: with open(txt_file, 'r', encoding='utf8') as f:
for line in f: for line in f:
cols = line.split('| ') cols = line.split('| ')
items.append(cols) # text, wav_full_path, mel_name, linear_name, wav_len, mel_len # text, wav_full_path, mel_name, linear_name, wav_len, mel_len
random.shuffle(items) items.append(cols)
return items return items
@ -23,10 +22,9 @@ def tweb(root_path, meta_file):
with open(txt_file, 'r') as ttf: with open(txt_file, 'r') as ttf:
for line in ttf: for line in ttf:
cols = line.split('\t') cols = line.split('\t')
wav_file = os.path.join(root_path, cols[0]+'.wav') wav_file = os.path.join(root_path, cols[0] + '.wav')
text = cols[1] text = cols[1]
items.append([text, wav_file]) items.append([text, wav_file])
random.shuffle(items)
return items return items
@ -44,48 +42,48 @@ def tweb(root_path, meta_file):
def mozilla(root_path, meta_files): def mozilla(root_path, meta_files):
"""Normalizes Mozilla meta data files to TTS format""" """Normalizes Mozilla meta data files to TTS format"""
import glob import glob
meta_files = glob.glob(root_path + "/**/batch*.txt", recursive=True) meta_files = glob.glob(root_path + "/**/batch*.txt", recursive=True)
folders = [os.path.dirname(f.strip()) for f in meta_files] folders = [os.path.dirname(f.strip()) for f in meta_files]
items = [] items = []
for idx, meta_file in enumerate(meta_files): for idx, meta_file in enumerate(meta_files):
folder = folders[idx] folder = folders[idx]
txt_file = os.path.join(root_path, meta_file) txt_file = os.path.join(root_path, meta_file)
with open(txt_file, 'r') as ttf: with open(txt_file, 'r') as ttf:
for line in ttf: for line in ttf:
cols = line.split('|') cols = line.split('|')
wav_file = os.path.join(root_path, folder, 'wavs_no_processing', cols[1].strip()) wav_file = os.path.join(root_path, folder, 'wavs_no_processing',
if os.path.isfile(wav_file): cols[1].strip())
text = cols[0].strip() if os.path.isfile(wav_file):
items.append([text, wav_file]) text = cols[0].strip()
else: items.append([text, wav_file])
print(" > Error: {}".format(cols)) else:
continue print(" > Error: {}".format(cols))
random.shuffle(items) continue
return items return items
def mailabs(root_path, meta_files): def mailabs(root_path, meta_files):
"""Normalizes M-AI-Labs meta data files to TTS format""" """Normalizes M-AI-Labs meta data files to TTS format"""
folders = [os.path.dirname(f.strip()) for f in meta_files.split(",")] folders = [os.path.dirname(f.strip()) for f in meta_files.split(",")]
meta_files = [f.strip() for f in meta_files.split(",")] meta_files = [f.strip() for f in meta_files.split(",")]
items = [] items = []
for idx, meta_file in enumerate(meta_files): for idx, meta_file in enumerate(meta_files):
print(" | > {}".format(meta_file)) print(" | > {}".format(meta_file))
folder = folders[idx] folder = folders[idx]
txt_file = os.path.join(root_path, meta_file) txt_file = os.path.join(root_path, meta_file)
with open(txt_file, 'r') as ttf: with open(txt_file, 'r') as ttf:
for line in ttf: for line in ttf:
cols = line.split('|') cols = line.split('|')
wav_file = os.path.join(root_path, folder, 'wavs', cols[0]+'.wav') wav_file = os.path.join(root_path, folder, 'wavs',
if os.path.isfile(wav_file): cols[0] + '.wav')
text = cols[1] if os.path.isfile(wav_file):
items.append([text, wav_file]) text = cols[1]
else: items.append([text, wav_file])
continue else:
random.shuffle(items) continue
return items return items
def ljspeech(root_path, meta_file): def ljspeech(root_path, meta_file):
@ -95,10 +93,9 @@ def ljspeech(root_path, meta_file):
with open(txt_file, 'r') as ttf: with open(txt_file, 'r') as ttf:
for line in ttf: for line in ttf:
cols = line.split('|') cols = line.split('|')
wav_file = os.path.join(root_path, 'wavs', cols[0]+'.wav') wav_file = os.path.join(root_path, 'wavs', cols[0] + '.wav')
text = cols[1] text = cols[1]
items.append([text, wav_file]) items.append([text, wav_file])
random.shuffle(items)
return items return items
@ -109,10 +106,9 @@ def nancy(root_path, meta_file):
with open(txt_file, 'r') as ttf: with open(txt_file, 'r') as ttf:
for line in ttf: for line in ttf:
id = line.split()[1] id = line.split()[1]
text = line[line.find('"')+1:line.rfind('"')-1] text = line[line.find('"') + 1:line.rfind('"') - 1]
wav_file = os.path.join(root_path, "wavn", id + ".wav") wav_file = os.path.join(root_path, "wavn", id + ".wav")
items.append([text, wav_file]) items.append([text, wav_file])
random.shuffle(items)
return items return items
@ -126,7 +122,6 @@ def common_voice(root_path, meta_file):
continue continue
cols = line.split("\t") cols = line.split("\t")
text = cols[2] text = cols[2]
# Files need to be first converted to wav...
wav_file = os.path.join(root_path, "clips", cols[1] + ".wav") wav_file = os.path.join(root_path, "clips", cols[1] + ".wav")
items.append([text, wav_file]) items.append([text, wav_file])
return items return items