i use jupyterlab to run the script seq2seq_with_attn.py (converted to .ipny)

in the function
sents_en, sents_fr_in, sents_fr_out = download_and_read(download_url, num_sent_pairs=NUM_SENT_PAIRS)

there is reported error
UnicodeDecodeError: 'cp950' codec can't decode byte 0xe2 in position 555: illegal multibyte sequence

i check on web and find that, it needs to add encoding="utf-8" for the line
with open(local_file, "r") as fin:

however, after done that, there is another bug reported in the line
en_sent, fr_sent = line.strip().split('\t')
ValueError: too many values to unpack (expected 2)
the first read line is as
['Go.',
'Va !',
'CC-BY 2.0 (France) Attribution: tatoeba.org #2877272 (CM) & #1158250 (Wittydev)']
so only two returned values en_sent, fr_sent are not enough

history = model.fit(x=train_features,y=train_labels, epochs=100, verbose=1, validation_split=0.2)

ValueError Traceback (most recent call last)
/root/repos/Deep-Learning-with-TensorFlow-and-Keras-3rd-edition/Chapter_2/multiple_linear_regression_using_keras_API.ipynb Cell 13 line 1
----> 1 history = model.fit(x=train_features,y=train_labels, epochs=100, verbose=1, validation_split=0.2)

File /usr/local/lib/python3.11/dist-packages/keras/src/utils/traceback_utils.py:70, in filter_traceback..error_handler(*args, **kwargs)
67 filtered_tb = _process_traceback_frames(e.traceback)
68 # To get the full stack trace, call:
69 # tf.debugging.disable_traceback_filtering()
---> 70 raise e.with_traceback(filtered_tb) from None
71 finally:
72 del filtered_tb

File /usr/local/lib/python3.11/dist-packages/keras/src/engine/data_adapter.py:1776, in train_validation_split(arrays, validation_split)
1774 unsplitable = [type(t) for t in flat_arrays if not _can_split(t)]
1775 if unsplitable:
-> 1776 raise ValueError(
1777 "validation_split is only supported for Tensors or NumPy "
1778 "arrays, found following types in the input: {}".format(unsplitable)
1779 )
1781 if all(t is None for t in flat_arrays):
1782 return arrays, arrays

ValueError: validation_split is only supported for Tensors or NumPy arrays, found following types in the input: [<class 'pandas.core.frame.DataFrame'>, <class 'pandas.core.series.Series'>]

OS: linux (docker container)
Python: 3.11.0rc1
keras: 2.14.0
Tensorflow: 2.14.0

spam_classifier.py contains the following code to evaluate the accuracy of the model:

evaluate against test set

labels, predictions = [], []
for Xtest, Ytest in test_dataset:
Ytest_ = model.predict_on_batch(Xtest)
ytest = np.argmax(Ytest, axis=1)
ytest_ = np.argmax(Ytest_, axis=1)
labels.extend(ytest.tolist())
predictions.extend(ytest.tolist())

print("test accuracy: {:.3f}".format(accuracy_score(labels, predictions)))
print("confusion matrix")
print(confusion_matrix(labels, predictions))

The accuracy, however, is always 100% as the two lists - labels and predictions - get the same list added to them - ytest. The following update should be made:

evaluate against test set

labels, predictions = [], []
for Xtest, Ytest in test_dataset:
Ytest_ = model.predict_on_batch(Xtest)
ytest = np.argmax(Ytest, axis=1)
ytest_ = np.argmax(Ytest_, axis=1)
labels.extend(ytest.tolist())
predictions.extend(ytest_.tolist())

print("test accuracy: {:.3f}".format(accuracy_score(labels, predictions)))
print("confusion matrix")
print(confusion_matrix(labels, predictions))

for the call function in the Decoder class,
the input to the self.attention should be (state, encoder_out)?

def call(self, x, state, encoder_out):
x = self.embedding(x)
context, alignment = self.attention(x, encoder_out)
x = tf.expand_dims(
tf.concat([
x, tf.squeeze(context, axis=1)
], axis=1),
axis=1)
x, state = self.rnn(x, state)
x = self.Wc(x)
x = self.Ws(x)
return x, state, alignment

V_ic_t-1 variable should not be present in the calculation of i, f, and o gates of LSTM. It is used in peephole LSTM.