malarinv
/
tacotron2
mirror of https://github.com/malarinv/tacotron2
1
0
Fork 0
Code Issues Releases Wiki Activity

Merge pull request #19 from NVIDIA/load_mel_from_disk 

Load mel from disk
fp16_path
Rafael Valle 2018-05-15 08:54:24 -07:00 committed by GitHub
parent 20568643cb 2da7a2ebab
commit bd42cb6ed7
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 17 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
README.md
View File

@ -20,6 +20,7 @@ Distributed and FP16 support relies on work by Christian Sarofeen and NVIDIA's
2. Clone this repo: `git clone https://github.com/NVIDIA/tacotron2.git` 2. Clone this repo: `git clone https://github.com/NVIDIA/tacotron2.git`
3. CD into this repo: `cd tacotron2` 3. CD into this repo: `cd tacotron2`
4. Update .wav paths: `sed -i -- 's,DUMMY,ljs_dataset_folder/wavs,g' filelists/*.txt` 4. Update .wav paths: `sed -i -- 's,DUMMY,ljs_dataset_folder/wavs,g' filelists/*.txt`
- Alternatively, set `load_mel_from_disk=True` in `hparams.py` and update mel-spectrogram paths
5. Install [pytorch 0.4](https://github.com/pytorch/pytorch) 5. Install [pytorch 0.4](https://github.com/pytorch/pytorch)
6. Install python requirements or build docker image 6. Install python requirements or build docker image
- Install python requirements: `pip install -r requirements.txt` - Install python requirements: `pip install -r requirements.txt`

9
data_utils.py
View File

@ -1,4 +1,5 @@
import random import random
import numpy as np
import torch import torch
import torch.utils.data import torch.utils.data
@ -19,6 +20,7 @@ class TextMelLoader(torch.utils.data.Dataset):
self.text_cleaners = hparams.text_cleaners self.text_cleaners = hparams.text_cleaners
self.max_wav_value = hparams.max_wav_value self.max_wav_value = hparams.max_wav_value
self.sampling_rate = hparams.sampling_rate self.sampling_rate = hparams.sampling_rate
self.load_mel_from_disk = hparams.load_mel_from_disk
self.stft = layers.TacotronSTFT( self.stft = layers.TacotronSTFT(
hparams.filter_length, hparams.hop_length, hparams.win_length, hparams.filter_length, hparams.hop_length, hparams.win_length,
hparams.n_mel_channels, hparams.sampling_rate, hparams.mel_fmin, hparams.n_mel_channels, hparams.sampling_rate, hparams.mel_fmin,
@ -35,12 +37,19 @@ class TextMelLoader(torch.utils.data.Dataset):
return (text, mel) return (text, mel)
def get_mel(self, filename): def get_mel(self, filename):
if not self.load_mel_from_disk:
audio = load_wav_to_torch(filename, self.sampling_rate) audio = load_wav_to_torch(filename, self.sampling_rate)
audio_norm = audio / self.max_wav_value audio_norm = audio / self.max_wav_value
audio_norm = audio_norm.unsqueeze(0) audio_norm = audio_norm.unsqueeze(0)
audio_norm = torch.autograd.Variable(audio_norm, requires_grad=False) audio_norm = torch.autograd.Variable(audio_norm, requires_grad=False)
melspec = self.stft.mel_spectrogram(audio_norm) melspec = self.stft.mel_spectrogram(audio_norm)
melspec = torch.squeeze(melspec, 0) melspec = torch.squeeze(melspec, 0)
else:
melspec = torch.from_numpy(np.load(filename))
assert melspec.size(0) == self.stft.n_mel_channels, (
'Mel dimension mismatch: given {}, expected {}'.format(
melspec.size(0), self.stft.n_mel_channels))
return melspec return melspec
def get_text(self, text): def get_text(self, text):

1
hparams.py
View File

@ -23,6 +23,7 @@ def create_hparams(hparams_string=None, verbose=False):
################################ ################################
# Data Parameters # # Data Parameters #
################################ ################################
load_mel_from_disk=False,
training_files='filelists/ljs_audio_text_train_filelist.txt', training_files='filelists/ljs_audio_text_train_filelist.txt',
validation_files='filelists/ljs_audio_text_val_filelist.txt', validation_files='filelists/ljs_audio_text_val_filelist.txt',
text_cleaners=['english_cleaners'], text_cleaners=['english_cleaners'],