Getting Value Error while creating RNN model - tensorflow

Question

Im getting this error: ValueError: logits and labels must have the same shape, received ((32, 1) vs (32, 23740))
This is my full code:
import pandas as pd
from keras.models import Sequential
from keras.layers import Dense, Embedding, LSTM
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix, accuracy_score, precision_score, recall_score, f1_score
import re
import numpy as np
from sklearn.preprocessing import StandardScaler, OneHotEncoder, LabelEncoder
from keras.preprocessing.text import Tokenizer
from keras.utils import to_categorical
from keras.utils import pad_sequences
# All data processing stuff
data = pd.read_csv('test.csv') # Load the data
# convert Sentiment type to string
data['Sentiment'] = data['Sentiment'].astype(str)
# remove special characters and convert to lowercase
data["Tweet"] = data["Tweet"].apply(lambda x: x.lower()) # Convert all tweets to lowercase
data["Tweet"] = data["Tweet"].apply(lambda x: x.replace("[^a-zA-Z0-9]", "")) # Remove special characters
data["Sentiment"] = data["Sentiment"].apply(lambda x: x.lower())
# Initialize the Tokenizer
tokenizer = Tokenizer()
# Fit the Tokenizer on the text data
tokenizer.fit_on_texts(data["Tweet"] + data["Sentiment"])
token_sequences1 = tokenizer.texts_to_matrix(data["Tweet"])
token_sequences2 = tokenizer.texts_to_matrix(data["Sentiment"])
padded_sequences1 = pad_sequences(token_sequences1)
padded_sequences2 = pad_sequences(token_sequences2)
# split the data into training and testing sets
x_train, x_test, y_train, y_test = train_test_split(
padded_sequences1, padded_sequences2, test_size=0.2, random_state=42)
# create the RNN model
model = Sequential()
# 1000 is the number of words in the vocabulary, 128 is the dimension of the embedding vector,
model.add(Embedding(1000, 128))
model.add(LSTM(128, dropout=0.2, recurrent_dropout=0.2))
model.add(Dense(1, activation='sigmoid'))
# compile the model with a specified loss function and optimizer
# binary_crossentropy is used for binary classification problems like this one (positive or negative sentiment)
# accuracy is the metric used to evaluate the model performance (the percentage of correct predictions)
# the loss function and the optimizer can be changed to see if the model performance improves or not
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
print(model.summary())
# train the model on the training data and put verbose=1 to see the training progress
model.fit(x_train, y_train, batch_size=32, epochs=10, verbose=1)
# evaluate the model on the testing data
y_pred = model.predict(x_test)
# calculate the evaluation metrics
accuracy = accuracy_score(y_test, y_pred)
precision = precision_score(y_test, y_pred)
recall = recall_score(y_test, y_pred)
f1 = f1_score(y_test, y_pred)
# print the evaluation metrics
print("Accuracy: {:.2f}".format(accuracy))
print("Precision: {:.2f}".format(precision))
print("Recall: {:.2f}".format(recall))
print("F1 Score: {:.2f}".format(f1))
Im actually trying to create sentiment analyser on news headlines tweets from this data, here is a sample image of how my data looks in CSV.
Please provide a solution to this, I have tried every solution I had find