mirror of https://github.com/coqui-ai/TTS.git
5.9 MiB
5.9 MiB
None
<html lang="en">
<head>
</head>
![No description has been provided for this image]()
![No description has been provided for this image]()
![No description has been provided for this image]()
![No description has been provided for this image]()
![No description has been provided for this image]()
![No description has been provided for this image]()
![No description has been provided for this image]()
![No description has been provided for this image]()
![No description has been provided for this image]()
![No description has been provided for this image]()
</html>
In [16]:
%load_ext autoreload %autoreload 2 import os import sys import io import torch import time import numpy as np from collections import OrderedDict from matplotlib import pylab as plt %pylab inline rcParams["figure.figsize"] = (16,5) sys.path.append('/home/erogol/projects/') import librosa import librosa.display from TTS.models.tacotron import Tacotron from TTS.layers import * from TTS.utils.data import * from TTS.utils.audio import AudioProcessor from TTS.utils.generic_utils import load_config from TTS.utils.text import text_to_sequence import IPython from IPython.display import Audio from utils import *
The autoreload extension is already loaded. To reload it, use: %reload_ext autoreload Populating the interactive namespace from numpy and matplotlib
/home/erogol/miniconda3/envs/pytorch4/lib/python3.6/site-packages/IPython/core/magics/pylab.py:160: UserWarning: pylab import has clobbered these variables: ['plt'] `%matplotlib` prevents importing * from pylab and numpy "\n`%matplotlib` prevents importing * from pylab and numpy"
In [17]:
ls /data/shared/erogol_models/May-14-2018_07:10PM-stop-token-separate-larger-lr-b2e09f9
best_model.pth.tar checkpoint_35344.pth.tar checkpoint_10152.pth.tar checkpoint_35720.pth.tar checkpoint_10528.pth.tar checkpoint_36096.pth.tar checkpoint_10904.pth.tar checkpoint_36472.pth.tar checkpoint_1128.pth.tar checkpoint_36848.pth.tar checkpoint_11280.pth.tar checkpoint_37224.pth.tar checkpoint_11656.pth.tar checkpoint_376.pth.tar checkpoint_12032.pth.tar checkpoint_3760.pth.tar checkpoint_12408.pth.tar checkpoint_37600.pth.tar checkpoint_12784.pth.tar checkpoint_37976.pth.tar checkpoint_13160.pth.tar checkpoint_38352.pth.tar checkpoint_13536.pth.tar checkpoint_38728.pth.tar checkpoint_13912.pth.tar checkpoint_39104.pth.tar checkpoint_14288.pth.tar checkpoint_39480.pth.tar checkpoint_14664.pth.tar checkpoint_39856.pth.tar checkpoint_1504.pth.tar checkpoint_40232.pth.tar checkpoint_15040.pth.tar checkpoint_40608.pth.tar checkpoint_15416.pth.tar checkpoint_40984.pth.tar checkpoint_15792.pth.tar checkpoint_4136.pth.tar checkpoint_16168.pth.tar checkpoint_41360.pth.tar checkpoint_16544.pth.tar checkpoint_41736.pth.tar checkpoint_16920.pth.tar checkpoint_42112.pth.tar checkpoint_17296.pth.tar checkpoint_42488.pth.tar checkpoint_17672.pth.tar checkpoint_42864.pth.tar checkpoint_18048.pth.tar checkpoint_43240.pth.tar checkpoint_18424.pth.tar checkpoint_43616.pth.tar checkpoint_1880.pth.tar checkpoint_43992.pth.tar checkpoint_18800.pth.tar checkpoint_44368.pth.tar checkpoint_19176.pth.tar checkpoint_44744.pth.tar checkpoint_19552.pth.tar checkpoint_4512.pth.tar checkpoint_19928.pth.tar checkpoint_45120.pth.tar checkpoint_20304.pth.tar checkpoint_45496.pth.tar checkpoint_20680.pth.tar checkpoint_45872.pth.tar checkpoint_21056.pth.tar checkpoint_46248.pth.tar checkpoint_21432.pth.tar checkpoint_46624.pth.tar checkpoint_21808.pth.tar checkpoint_47000.pth.tar checkpoint_22184.pth.tar checkpoint_47376.pth.tar checkpoint_2256.pth.tar checkpoint_47752.pth.tar checkpoint_22560.pth.tar checkpoint_48128.pth.tar checkpoint_22936.pth.tar checkpoint_48504.pth.tar checkpoint_23312.pth.tar checkpoint_4888.pth.tar checkpoint_23688.pth.tar checkpoint_48880.pth.tar checkpoint_24064.pth.tar checkpoint_49256.pth.tar checkpoint_24440.pth.tar checkpoint_49632.pth.tar checkpoint_24816.pth.tar checkpoint_50008.pth.tar checkpoint_25192.pth.tar checkpoint_50384.pth.tar checkpoint_25568.pth.tar checkpoint_50760.pth.tar checkpoint_25944.pth.tar checkpoint_51136.pth.tar checkpoint_2632.pth.tar checkpoint_51512.pth.tar checkpoint_26320.pth.tar checkpoint_51888.pth.tar checkpoint_26696.pth.tar checkpoint_52264.pth.tar checkpoint_27072.pth.tar checkpoint_5264.pth.tar checkpoint_27448.pth.tar checkpoint_52640.pth.tar checkpoint_27824.pth.tar checkpoint_53016.pth.tar checkpoint_28200.pth.tar checkpoint_53392.pth.tar checkpoint_28576.pth.tar checkpoint_53768.pth.tar checkpoint_28952.pth.tar checkpoint_54144.pth.tar checkpoint_29328.pth.tar checkpoint_54520.pth.tar checkpoint_29704.pth.tar checkpoint_54896.pth.tar checkpoint_3008.pth.tar checkpoint_5640.pth.tar checkpoint_30080.pth.tar checkpoint_6016.pth.tar checkpoint_30456.pth.tar checkpoint_6392.pth.tar checkpoint_30832.pth.tar checkpoint_6768.pth.tar checkpoint_31208.pth.tar checkpoint_7144.pth.tar checkpoint_31584.pth.tar checkpoint_752.pth.tar checkpoint_31960.pth.tar checkpoint_7520.pth.tar checkpoint_32336.pth.tar checkpoint_7896.pth.tar checkpoint_32712.pth.tar checkpoint_8272.pth.tar checkpoint_33088.pth.tar checkpoint_8648.pth.tar checkpoint_33464.pth.tar checkpoint_9024.pth.tar checkpoint_3384.pth.tar checkpoint_9400.pth.tar checkpoint_33840.pth.tar checkpoint_9776.pth.tar checkpoint_34216.pth.tar checkpoints/ checkpoint_34592.pth.tar config.json checkpoint_34968.pth.tar events.out.tfevents.1526350247.mlc1
In [18]:
def tts(model, text, CONFIG, use_cuda, ap, figures=True): t_1 = time.time() waveform, alignment, spectrogram, stop_tokens = create_speech(model, text, CONFIG, use_cuda, ap) print(" > Run-time: {}".format(time.time() - t_1)) if figures: visualize(alignment, spectrogram, stop_tokens, CONFIG) IPython.display.display(Audio(waveform, rate=CONFIG.sample_rate)) return alignment, spectrogram, stop_tokens
In [19]:
# Set constants ROOT_PATH = '/data/shared/erogol_models/May-15-2018_04:12PM-stop-token-rnn-40f1a3d/' MODEL_PATH = ROOT_PATH + '/checkpoint_16920.pth.tar' CONFIG_PATH = ROOT_PATH + '/config.json' OUT_FOLDER = ROOT_PATH + '/test/' CONFIG = load_config(CONFIG_PATH) use_cuda = True
In [20]:
# load the model model = Tacotron(CONFIG.embedding_size, CONFIG.num_freq, CONFIG.num_mels, CONFIG.r) # load the audio processor ap = AudioProcessor(CONFIG.sample_rate, CONFIG.num_mels, CONFIG.min_level_db, CONFIG.frame_shift_ms, CONFIG.frame_length_ms, CONFIG.preemphasis, CONFIG.ref_level_db, CONFIG.num_freq, CONFIG.power, griffin_lim_iters=30) # load model state if use_cuda: cp = torch.load(MODEL_PATH) else: cp = torch.load(MODEL_PATH, map_location=lambda storage, loc: storage) # load the model model.load_state_dict(cp['model']) if use_cuda: model.cuda() model.eval()
| > Number of characters : 149
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-20-33e7f6a6cc7d> in <module>() 1 # load the model ----> 2 model = Tacotron(CONFIG.embedding_size, CONFIG.num_freq, CONFIG.num_mels, CONFIG.r) 3 4 # load the audio processor 5 ~/projects/TTS/models/tacotron.py in __init__(self, embedding_dim, linear_dim, mel_dim, r, padding_idx) 17 print(" | > Number of characters : {}".format(len(symbols))) 18 self.embedding.weight.data.normal_(0, 0.3) ---> 19 self.encoder = Encoder(embedding_dim) 20 self.decoder = Decoder(256, mel_dim, r) 21 self.postnet = CBHG(mel_dim, K=8, projections=[256, mel_dim]) ~/projects/TTS/layers/tacotron.py in __init__(self, in_features) 188 189 def __init__(self, in_features): --> 190 super(Encoder, self).__init__() 191 self.prenet = Prenet(in_features, out_features=[256, 128]) 192 self.cbhg = CBHG(128, K=16, projections=[128, 128]) TypeError: super(type, obj): obj must be an instance or subtype of type
EXAMPLES FROM TRAINING SET¶
In [21]:
import pandas as pd df = pd.read_csv('/data/shared/KeithIto/LJSpeech-1.0/metadata_val.csv', delimiter='|')
In [22]:
sentence = df.iloc[175, 1] print(sentence) model.decoder.max_decoder_steps = 250 align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)
Hosty took the necessary steps to have the Dallas office of the FBI, rather than the New Orleans office, reestablished as the office with principal responsibility.
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 6.3198254108428955
Your browser does not support the audio element.
Comparision with https://mycroft.ai/blog/available-voices/¶
In [23]:
sentence = "It took me quite a long time to develop a voice, and now that I have it I'm not going to be silent." model.decoder.max_decoder_steps = 250 align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap, figures=True)
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 4.629634857177734
Your browser does not support the audio element.
In [24]:
sentence = "Be a voice,not an echo." # 'echo' is not in training set. align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 1.2954251766204834
Your browser does not support the audio element.
In [25]:
sentence = "The human voice is the most perfect instrument of all." align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 2.442563772201538
Your browser does not support the audio element.
In [26]:
sentence = "I'm sorry Dave. I'm afraid I can't do that." align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 2.959503412246704
Your browser does not support the audio element.
In [27]:
sentence = "This cake is great. It's so delicious and moist." align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 3.3856382369995117
Your browser does not support the audio element.
Comparison with https://keithito.github.io/audio-samples/¶
In [28]:
sentence = "Generative adversarial network or variational auto-encoder." align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 3.539504289627075
Your browser does not support the audio element.
In [29]:
sentence = "Scientists at the CERN laboratory say they have discovered a new particle." align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 3.4460177421569824
Your browser does not support the audio element.
In [30]:
sentence = "here’s a way to measure the acute emotional intelligence that has never gone out of style." align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 3.5448789596557617
Your browser does not support the audio element.
In [31]:
sentence = "President Trump met with other leaders at the Group of 20 conference." align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)
/home/erogol/projects/TTS/notebooks/utils.py:19: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead. torch.from_numpy(seq), volatile=True).unsqueeze(0).cuda()
> Run-time: 2.8492112159729004
Your browser does not support the audio element.
In [ ]:
sentence = "The buses aren't the problem, they actually provide a solution." align, spec, stop_tokens = tts(model, sentence, CONFIG, use_cuda, ap)