안녕하세요.

타코트론에 관심이 많아서(개발자는 아니지만 ^^;;) 하나씩 따라하다 보니 마지막 python test2.py 을 실행하고 나면

Traceback (most recent call last):
File "test2.py", line 42, in
test_step(mel, i)
File "test2.py", line 31, in test_step
scipy.io.wavfile.write(os.path.join(save_dir, '{}.wav'.format(idx)), sample_rate, wav)
AttributeError: module 'scipy' has no attribute 'io'

이러한 에러가 발생을 합니다.

텐스플로우는 최신버전이 설치되어 있습니다. (2.0 이상)

Traceback (most recent call last):
File "preprocess.py", line 28, in
sentence = text_to_sequence(sentence)
File "F:\Tacotron-Korean-Tensorflow2-master\util\text.py", line 23, in text_to_sequence
sequence.append(_symbol_to_id[s])
KeyError: '3'

위 같은 에러가
python preprocess.py 를 칠때 나타나는데 어떻해야하나요?

When I run train1.py I got error message as below. Can you teach me how to resolve it? Thank you.
p.s. I am running this project on Windows10

2021-03-23 23:00:28.775696: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
WARNING:tensorflow:From C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\autograph\impl\api.py:330: bidirectional_dynamic_rnn (from tensorflow.python.ops.rnn) is deprecated and will be removed in a future version.
Instructions for updating:
Please use keras.layers.Bidirectional(keras.layers.RNN(cell)), which is equivalent to this API
WARNING:tensorflow:From C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\ops\rnn.py:464: dynamic_rnn (from tensorflow.python.ops.rnn) is deprecated and will be removed in a future version.
Instructions for updating:
Please use keras.layers.RNN(cell), which is equivalent to this API
Traceback (most recent call last):
File "train1.py", line 88, in
loss, pred, alignment = train_step(text, dec, mel, text_length)
File "C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\eager\def_function.py", line 457, in call
result = self._call(*args, **kwds)
File "C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\eager\def_function.py", line 503, in _call
self._initialize(args, kwds, add_initializers_to=initializer_map)
File "C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\eager\def_function.py", line 408, in _initialize
*args, **kwds))
File "C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\eager\function.py", line 1848, in _get_concrete_function_internal_garbage_collected
graph_function, _, _ = self._maybe_define_function(args, kwargs)
File "C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\eager\function.py", line 2150, in _maybe_define_function
graph_function = self._create_graph_function(args, kwargs)
File "C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\eager\function.py", line 2041, in _create_graph_function
capture_by_value=self._capture_by_value),
File "C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\framework\func_graph.py", line 915, in func_graph_from_py_func
func_outputs = python_func(*func_args, **func_kwargs)
File "C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\eager\def_function.py", line 358, in wrapped_fn
return weak_wrapped_fn().wrapped(*args, **kwds)
File "C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\framework\func_graph.py", line 905, in wrapper
raise e.ag_error_metadata.to_exception(e)
NotImplementedError: in converted code:

train1.py:57 train_step  *
    pred, alignment = model(enc_input, text_length, dec_input, is_training=True)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\engine\base_layer.py:847 __call__
    outputs = call_fn(cast_inputs, *args, **kwargs)
D:\GitHUB\Tacotron-Korean-Tensorflow2\models\tacotron.py:54 call  *
    x = self.decoder(batch, dec_input, x)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\engine\base_layer.py:891 __call__
    outputs = self.call(cast_inputs, *args, **kwargs)
D:\GitHUB\Tacotron-Korean-Tensorflow2\models\tacotron.py:32 call
    x = self.attention_rnn(x)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\layers\recurrent.py:623 __call__
    return super(RNN, self).__call__(inputs, **kwargs)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\engine\base_layer.py:847 __call__
    outputs = call_fn(cast_inputs, *args, **kwargs)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\layers\recurrent_v2.py:313 call
    inputs, initial_state, _ = self._process_inputs(inputs, initial_state, None)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\layers\recurrent.py:798 _process_inputs
    initial_state = self.get_initial_state(inputs)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\layers\recurrent.py:606 get_initial_state
    inputs=None, batch_size=batch_size, dtype=dtype)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\layers\recurrent.py:1762 get_initial_state
    return _generate_zero_filled_state_for_cell(self, inputs, batch_size, dtype)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\layers\recurrent.py:2752 _generate_zero_filled_state_for_cell
    return _generate_zero_filled_state(batch_size, cell.state_size, dtype)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\layers\recurrent.py:2770 _generate_zero_filled_state
    return create_zeros(state_size)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\keras\layers\recurrent.py:2765 create_zeros
    return array_ops.zeros(init_state_size, dtype=dtype)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\ops\array_ops.py:2349 zeros
    output = _constant_if_small(zero, shape, dtype, name)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\ops\array_ops.py:2306 _constant_if_small
    if np.prod(shape) < 1000:
<__array_function__ internals>:6 prod

C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\numpy\core\fromnumeric.py:3031 prod
    keepdims=keepdims, initial=initial, where=where)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\numpy\core\fromnumeric.py:87 _wrapreduction
    return ufunc.reduce(obj, axis, dtype, out, **passkwargs)
C:\Users\lee_j\Anaconda3\envs\Lecture\lib\site-packages\tensorflow_core\python\framework\ops.py:736 __array__
    " array.".format(self.name))

NotImplementedError: Cannot convert a symbolic Tensor (tacotron/decoder/gru/strided_slice:0) to a numpy array.

Hi. How long does the training module typically take? Is there a maximum defined step-count after which the training will converge? How to utilize my gpu to run this code? should I change any configuration parameters if I use my own dataset?