Sequence to sequence input dimension error

Question

from re import search
import numpy as np
import tensorflow as tf
import pandas as pd
from tensorflow.keras.layers import Dense,LSTM,Embedding,Input,Bidirectional
from tensorflow.keras.models import Model
from tensorflow.keras.preprocessing.text import Tokenizer
from tensorflow.keras.preprocessing.sequence import pad_sequences


df=pd.read_excel(r"C:/Users/Shubham/Downloads/pdf.xlsx")

df.drop('Unnamed: 0',axis=1,inplace=True)

df.columns=['incorrect','correct']

df['incorrect']=df['incorrect'].apply(lambda x: str(x))
df['correct']=df['correct'].apply(lambda x: str(x))

train=df[:1000]
test=df[1000:1100]
cv=df[1100:]

train['correct_inp']="$"+train['correct'].astype(str)
train['correct_out']=train['correct'].astype(str)+"@"

cv['correct_inp']="$"+cv['correct'].astype(str)
cv['correct_out']=cv['correct'].astype(str)+"@"

# tokenizing encoder input

tokenizer_incorr = Tokenizer(filters="",lower=False)
tokenizer_incorr.fit_on_texts(train["incorrect"].values)

incorr_train = np.array(tokenizer_incorr.texts_to_sequences(train["incorrect"].values))
incorr_cv = np.array(tokenizer_incorr.texts_to_sequences(cv["incorrect"].values))
print("vocab size of incorrrect sentences is",len(tokenizer_incorr.word_index))

# tokenizing decoder input

tokenizer_corr_inp = Tokenizer(filters="",lower=False)
tokenizer_corr_inp.fit_on_texts(train["correct_inp"].values)

corr_train_inp = np.array(tokenizer_corr_inp.texts_to_sequences(train["correct_inp"].values))
corr_cv_inp = np.array(tokenizer_corr_inp.texts_to_sequences(cv["correct_inp"].values))
print("vocab size of corrrect sentences is",len(tokenizer_corr_inp.word_index))

# tokenizing decoder output

tokenizer_corr_out = Tokenizer(filters="",lower=False)
tokenizer_corr_out.fit_on_texts(train["correct_out"].values)

corr_train_out = np.array(tokenizer_corr_out.texts_to_sequences(train["correct_out"].values))
corr_cv_out = np.array(tokenizer_corr_inp.texts_to_sequences(cv["correct_out"].values))

# padding training data
# here we are keeping max length a 25, as we ecided from EDA
incorr_train = 
np.array(pad_sequences(incorr_train,maxlen=160,padding="post",truncating='post'))
corr_train_inp = 
np.array(pad_sequences(corr_train_inp,maxlen=160,padding="post",truncating='post'))
corr_train_out = 
np.array(pad_sequences(corr_train_out,maxlen=160,padding="post",truncating='post'))


# padding validation data
incorr_cv = np.array(pad_sequences(incorr_cv,maxlen=160,padding="post",truncating='post'))
corr_cv_inp = np.array(pad_sequences(corr_cv_inp,maxlen=160,padding="post",truncating='post'))
corr_cv_out = np.array(pad_sequences(corr_cv_out,maxlen=160,padding="post",truncating='post'))


encoder_inputs = Input(shape=(160,))
encoder_embedding = Embedding(len(tokenizer_incorr.word_index)+1,128)(encoder_inputs)
encoder_lstm=Bidirectional(LSTM(256,return_sequences=True,return_state=True)) 
(encoder_embedding)
encoder_outputs, fw_state_h, fw_state_c, bw_state_h, bw_state_c =encoder_lstm

decoder_inputs=Input(shape=(160,))
decoder_embedding=Embedding(len(tokenizer_incorr.word_index)+1,128)(decoder_inputs)
decoder_lstm=Bidirectional(LSTM(256,return_sequences=True,return_state=True)) 
(decoder_embedding,initial_state=[fw_state_h, fw_state_c, bw_state_h, bw_state_c])
decoder_outputs=decoder_lstm
decoder_outputs=Dense(len(tokenizer_corr_inp.word_index)+1,activation='softmax') 

(decoder_outputs) model=Model([encoder_inputs,decoder_inputs],decoder_outputs) model.compile(optimizer='adam',loss='sparse_categorical_crossentropy',metrics=['accuracy'])

model.fit([incorr_train,corr_train_inp],corr_train_out,batch_size=32,epochs=10,
validation_data=([incorr_cv,corr_cv_inp],corr_cv_out))

This is the error i am getting right now. Please help

ValueError: Layer "dense_5" expects 1 input(s), but it received 5 input tensors. Inputs received: [<tf.Tensor 'Placeholder:0' shape=(None, 160, 512) dtype=float32>, <tf.Tensor 'Placeholder_1:0' shape=(None, 256) dtype=float32>, <tf.Tensor 'Placeholder_2:0' shape=(None, 256) dtype=float32>, <tf.Tensor 'Placeholder_3:0' shape=(None, 256) dtype=float32>, <tf.Tensor 'Placeholder_4:0' shape=(None, 256) dtype=float32>]