Error while model optimization using stm32 AI developer cloud

[Ritesh1]

 

I have created a tensorflow model for image classification. But during the optimization process using STM32 AI developer cloud, I am getting the following error:

error: Error when deserializing class 'InputLayer' using config={'batch_shape': [None, 32, 32, 3], 'dtype': 'float32', 'sparse': False, 'name': 'input_layer'}. Exception encountered: Unrecognized keyword arguments: ['batch_shape']

 

model = models.Sequential([
layers.Input(shape=(32, 32, 3)),
layers.Conv2D(32, (3, 3), activation='relu'),
layers.MaxPooling2D((2, 2)),
layers.Conv2D(64, (3, 3), activation='relu'),
layers.MaxPooling2D((2, 2)),
layers.Conv2D(64, (3, 3), activation='relu'),

layers.Flatten(),
layers.Dense(64, activation='relu'),
layers.Dense(10, activation='softmax') 
])

 

 

 

 

 

Ritesh

[Julian E.]

Hello @Ritesh1 ,

 

I don't know what you've done, but with a code looking like this, it works on st edge ai dev cloud:

 

import tensorflow as tf
from tensorflow.keras import layers, models

# Create the model
model = models.Sequential([
 layers.Input(shape=(32, 32, 3)), # Input layer with shape (32, 32, 3)
 layers.Conv2D(32, (3, 3), activation='relu'), # First convolutional layer with 32 filters, kernel size (3, 3), and ReLU activation
 layers.MaxPooling2D((2, 2)), # First max pooling layer with pool size (2, 2)
 layers.Conv2D(64, (3, 3), activation='relu'), # Second convolutional layer with 64 filters, kernel size (3, 3), and ReLU activation
 layers.MaxPooling2D((2, 2)), # Second max pooling layer with pool size (2, 2)
 layers.Conv2D(64, (3, 3), activation='relu'), # Third convolutional layer with 64 filters, kernel size (3, 3), and ReLU activation
 layers.Flatten(), # Flatten layer to convert 2D matrix to 1D vector
 layers.Dense(64, activation='relu'), # Fully connected layer with 64 units and ReLU activation
 layers.Dense(10, activation='softmax') # Output layer with 10 units (for classification) and softmax activation
])

# Compile the model
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])

# Save the model as an H5 file
model.save('model.h5')

# Summary of the model
model.summary()

 

Have a good day,

Julian