Import trained LSTM model on Streamlit

☁️ Community Cloud
1

I have trained a LSTM model on Colab and saved the model. Below is the code of my model:

from keras.layers import Dense, Dropout, LSTM
from keras.models import Sequential

# Build LSTM model
model = Sequential()  # Initialize the Sequential model
model.add(LSTM(units=50, activation='relu', return_sequences=True, input_shape=(x_train.shape[1], 1)))
model.add(Dropout(0.2))



model.add(LSTM(units = 60, activation = 'relu', return_sequences = 'True'))
model.add(Dropout(0.3))

model.add(LSTM(units = 80, activation = 'relu', return_sequences = 'True'))
model.add(Dropout(0.4))

model.add(LSTM(units = 120, activation = 'relu'))
model.add(Dropout(0.5))


model.add(Dense(units = 1))

# Fit the model on traing data
model.compile(optimizer='adam', loss = 'mean_squared_error')
model.fit(x_train, y_train, epochs = 100)

# Save the model
model.save('LSTM_model.keras')

Then, I load the model on Streamlit with the below code:

from tensorflow.keras.models import load_model

# Load the model
model = load_model('LSTM_model.keras')

The trained model LSTM_model.keras and the streamlit app named main.py both are kept in same directory/path. I run the straemlit app by executing the below command from the same directory where the two files are kept.

streamlit run main.py

But when I run the Streamlit app, I get the error:

ValueError: Could not interpret initializer identifier: {‘module’: ‘keras.initializers’, ‘class_name’: ‘Orthogonal’, ‘config’: {‘gain’: 1.0, ‘seed’: None}, ‘registered_name’: None, ‘shared_object_id’: 140299714528304}
Traceback:
File “D:\Program_Files\Python312\Lib\site-packages\streamlit\runtime\scriptrunner\script_runner.py”, line 542, in run_script
exec(code, module.dict)
File “C:\Users\sanni\Desktop\Stock\main.py”, line 72, in
model = load_model(‘LSTM_model.keras’)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “D:\Program_Files\Python312\Lib\site-packages\keras\src\saving\saving_api.py”, line 176, in load_model
return saving_lib.load_model(
^^^^^^^^^^^^^^^^^^^^^^
File “D:\Program_Files\Python312\Lib\site-packages\keras\src\saving\saving_lib.py”, line 155, in load_model
model = deserialize_keras_object(
^^^^^^^^^^^^^^^^^^^^^^^^^
File “D:\Program_Files\Python312\Lib\site-packages\keras\src\saving\serialization_lib.py”, line 711, in deserialize_keras_object
instance = cls.from_config(inner_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “D:\Program_Files\Python312\Lib\site-packages\keras\src\models\sequential.py”, line 331, in from_config
layer = serialization_lib.deserialize_keras_object(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “D:\Program_Files\Python312\Lib\site-packages\keras\src\saving\serialization_lib.py”, line 711, in deserialize_keras_object
instance = cls.from_config(inner_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File “D:\Program_Files\Python312\Lib\site-packages\keras\src\layers\rnn\lstm.py”, line 646, in from_config
return cls(**config)
^^^^^^^^^^^^^
File “D:\Program_Files\Python312\Lib\site-packages\keras\src\layers\rnn\lstm.py”, line 459, in init
cell = LSTMCell(
^^^^^^^^^
File “D:\Program_Files\Python312\Lib\site-packages\keras\src\layers\rnn\lstm.py”, line 122, in init
self.recurrent_initializer = initializers.get(recurrent_initializer)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\Program_Files\Python312\Lib\site-packages\keras\src\initializers_init.py", line 117, in get
raise ValueError(
2

The issue has been solved. The Google Colab use older version of Keras and Tensorflow, whereas in my PC has latest version of Keras and Tensorflow. I build and saved the model in Jupyter Notebook in my localhost, and it works as it use the latest version of Keras and Tensorflow installed in my PC since I am using localhost. I think the error occurred due to version mismatch.

1 Like