File size: 2,967 Bytes
d5ee97c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 |
# -*- coding: utf-8 -*-
# Copyright 2020 TensorFlowTTS Team.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import logging
import os
import pytest
import tensorflow as tf
from tensorflow_tts.configs import (
ParallelWaveGANGeneratorConfig,
ParallelWaveGANDiscriminatorConfig,
)
from tensorflow_tts.models import (
TFParallelWaveGANGenerator,
TFParallelWaveGANDiscriminator,
)
os.environ["CUDA_VISIBLE_DEVICES"] = ""
logging.basicConfig(
level=logging.DEBUG,
format="%(asctime)s (%(module)s:%(lineno)d) %(levelname)s: %(message)s",
)
def make_pwgan_generator_args(**kwargs):
defaults = dict(
out_channels=1,
kernel_size=3,
n_layers=30,
stacks=3,
residual_channels=64,
gate_channels=128,
skip_channels=64,
aux_channels=80,
aux_context_window=2,
dropout_rate=0.0,
use_bias=True,
use_causal_conv=False,
upsample_conditional_features=True,
upsample_params={"upsample_scales": [4, 4, 4, 4]},
initializer_seed=42,
)
defaults.update(kwargs)
return defaults
def make_pwgan_discriminator_args(**kwargs):
defaults = dict(
out_channels=1,
kernel_size=3,
n_layers=10,
conv_channels=64,
use_bias=True,
dilation_factor=1,
nonlinear_activation="LeakyReLU",
nonlinear_activation_params={"alpha": 0.2},
initializer_seed=42,
apply_sigmoid_at_last=False,
)
defaults.update(kwargs)
return defaults
@pytest.mark.parametrize(
"dict_g, dict_d",
[
({}, {}),
(
{"kernel_size": 3, "aux_context_window": 5, "residual_channels": 128},
{"dilation_factor": 2},
),
({"stacks": 4, "n_layers": 40}, {"conv_channels": 128}),
],
)
def test_melgan_trainable(dict_g, dict_d):
random_c = tf.random.uniform(shape=[4, 32, 80], dtype=tf.float32)
args_g = make_pwgan_generator_args(**dict_g)
args_d = make_pwgan_discriminator_args(**dict_d)
args_g = ParallelWaveGANGeneratorConfig(**args_g)
args_d = ParallelWaveGANDiscriminatorConfig(**args_d)
generator = TFParallelWaveGANGenerator(args_g)
generator._build()
discriminator = TFParallelWaveGANDiscriminator(args_d)
discriminator._build()
generated_audios = generator(random_c, training=True)
discriminator(generated_audios)
generator.summary()
discriminator.summary()
|