无法生成正确的音频

[EmptyMoon-hub]

你好，我运行了你的代码，将tacotron_batch_size改成了48，outputs_per_step改为3，但是输出在80000代时仍未对齐；
同时，我在训练仍未结束时测试synthesize，得到的确是1:14时常的wav，里面几乎全是杂音，这是错误提示

Loading checkpoint: logs-Tacotron-2/taco_pretrained/tacotron_model.ckpt-80000
WARNING:tensorflow:From /home/lyj/code/style-token-chinese/tacotron/synthesizer.py:63: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.
2020-10-07 17:29:27.822125: W tensorflow/compiler/jit/mark_for_compilation_pass.cc:1412] (One-time warning): Not using XLA:CPU for cluster because envvar TF_XLA_FLAGS=--tf_xla_cpu_global_jit was not set. If you want XLA:CPU, either set that envvar, or use experimental_jit_scope to enable XLA:CPU. To confirm that XLA is active, pass --vmodule=xla_compilation_cache=1 (as a proper command-line flag, not via TF_XLA_FLAGS) or set the envvar XLA_FLAGS=--xla_hlo_profile.
WARNING:tensorflow:From /home/lyj/code/style-token-chinese/tacotron/synthesizer.py:70: The name tf.train.Saver is deprecated. Please use tf.compat.v1.train.Saver instead.
WARNING:tensorflow:From /home/lyj/anaconda3/envs/my_Tensorflow_1/lib/python3.7/site-packages/tensorflow/python/training/saver.py:1276: checkpoint_exists (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.
Instructions for updating:
Use standard file APIs to check for files with this prefix.
Starting Synthesis
0%| | 0/1 [00:00<?, ?it/s]/home/lyj/anaconda3/envs/my_Tensorflow_1/lib/python3.7/site-packages/librosa/core/audio.py:162: UserWarning: PySoundFile failed. Trying audioread instead.
warnings.warn("PySoundFile failed. Trying audioread instead.")
/home/lyj/anaconda3/envs/my_Tensorflow_1/lib/python3.7/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 65292 missing from current font.
font.set_text(s, 0.0, flags=flags)
/home/lyj/anaconda3/envs/my_Tensorflow_1/lib/python3.7/site-packages/matplotlib/backends/backend_agg.py:238: RuntimeWarning: Glyph 12290 missing from current font.
font.set_text(s, 0.0, flags=flags)
/home/lyj/anaconda3/envs/my_Tensorflow_1/lib/python3.7/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 65292 missing from current font.
font.set_text(s, 0, flags=flags)
/home/lyj/anaconda3/envs/my_Tensorflow_1/lib/python3.7/site-packages/matplotlib/backends/backend_agg.py:201: RuntimeWarning: Glyph 12290 missing from current font.
font.set_text(s, 0, flags=flags)
100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [03:13<00:00, 193.99s/it]
synthesized mel spectrograms at tacotron_output/eval
Synthesizing audio from mel-spectrograms.. (This may take a while)
Traceback (most recent call last):
File "/home/lyj/code/style-token-chinese/wavenet_vocoder/synthesize.py", line 73, in wavenet_synthesize
checkpoint_path = tf.train.get_checkpoint_state(checkpoint).model_checkpoint_path
AttributeError: 'NoneType' object has no attribute 'model_checkpoint_path'
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "synthesize.py", line 100, in
main()
File "synthesize.py", line 94, in main
synthesize(args, hparams, taco_checkpoint, wave_checkpoint, sentences)
File "synthesize.py", line 42, in synthesize
wavenet_synthesize(args, hparams, wave_checkpoint)
File "/home/lyj/code/style-token-chinese/wavenet_vocoder/synthesize.py", line 76, in wavenet_synthesize
raise RuntimeError('Failed to load checkpoint at {}'.format(checkpoint))
RuntimeError: Failed to load checkpoint at logs-Tacotron-2/wave_pretrained/

推测是不是librosa这个库版本不对啊，我现在使用的是0.8.0

除此之外，我在阅读代码时并未找到哪里使用了pypinyin库处理中文数据集啊？貌似只是在utils里面存在，而在tacotron/utils里面并未使用pypinyin，请问这是怎么回事啊

[cnlinxi]

@EmptyMoon-hub

1. librosa库中PySoundFile没有调用成功，但这个应该不影响结果。
2. 需要提前利用pypinyin准备好数据集。输入到模型进行训练时，就应该是拼音形式了，而不能是汉字。
3. 你的错误是没有找到训练的ckpt模型文件。