LSTM 3D input 3D output dimension mismatch - keras

Question

I have a simple LSTM network in Keras:
def create_model(x):
lstm_model = Sequential()
lstm_model.add(LSTM(100, input_dim=x.shape[2], input_length=x.shape[1]))
lstm_model.add(Dense(1,activation='sigmoid'))
lstm_model.compile(loss='mean_squared_error', optimizer='adam')
return lstm_model
and I am trying to train it on the data that has the following shapes:
training data input: (100, 2784, 6), training data output: (100, 2784, 1)
validation data input: (50, 27, 6), validation data output: (50, 27, 1)
test data input: (50, 27, 6), test data output: (50, 27, 1)
model.fit(train_x, train_y, validation_data=(validation_x, validation_y), epochs=EPOCHS, batch_size=BATCH_SIZE, shuffle=False, callbacks=[...])
I keep on failing on dimensions either because a) validation data is not the same shape as the training data or b) because the y shapes are not right
What am I doing wrong?
P.S. Standalone code for convenience
import numpy as np
import keras
from keras.models import Sequential
from keras.layers import LSTM, Dense
x_train=np.arange(1670400)
x_train=x_train.reshape((100, 2784, 6))
y_train=np.arange(278400)
y_train=y_train.reshape((100, 2784, 1))
x_val=np.arange(8100)
x_val=x_val.reshape((50, 27, 6))
y_val=np.arange(1350)
y_val=y_val.reshape((50, 27, 1))
x_test=np.arange(8100)
x_test=x_test.reshape((50, 27, 6))
y_test=np.arange(1350)
y_test=y_test.reshape((50, 27, 1))
def create_model(x):
lstm_model = Sequential()
lstm_model.add(LSTM(100, input_dim=x.shape[2], input_length=x.shape[1]))
lstm_model.add(Dense(1,activation='sigmoid'))
lstm_model.compile(loss='mean_squared_error', optimizer='adam')
return lstm_model
model=create_model(x_train)
model.fit(x_train, y_train, validation_data=(x_val, y_val), epochs=10, batch_size=32, shuffle=False)

Answer 1

This modification worked:
changing to batch_input_shape and adding return_sequences=True to the LSTM layer
import numpy as np
import keras
from keras.models import Sequential
from keras.layers import LSTM, Dense
x_train=np.arange(1670400)
x_train=x_train.reshape((100, 2784, 6))
y_train=np.arange(278400)
y_train=y_train.reshape((100, 2784, 1))
x_val=np.arange(8100)
x_val=x_val.reshape((50, 27, 6))
y_val=np.arange(1350)
y_val=y_val.reshape((50, 27, 1))
x_test=np.arange(8100)
x_test=x_test.reshape((50, 27, 6))
y_test=np.arange(1350)
y_test=y_test.reshape((50, 27, 1))
def create_model():
lstm_model = Sequential()
lstm_model.add(LSTM(100, batch_input_shape=(None,None,6), return_sequences=True))
lstm_model.add(Dense(1, activation='sigmoid'))
lstm_model.compile(loss='mean_squared_error', optimizer='adam')
return lstm_model
model=create_model()
model.fit(x_train, y_train, validation_data=(x_val, y_val), epochs=10, batch_size=32, shuffle=False)