Error regarding grad cam implementation in model using tensorflow

[Senura Perera]

Hi I have an error regarding adding grad cam to my model.

this is the model summary 

tensor flow version is 2.13.0

this is the grad cam code

base_model = VGG16(weights='imagenet', include_top=False)

model = Sequential([base_model, Dense(5, activation='softmax')])

def visualize_grad_cam(input_model, img_array, layer_name, class_index, colormap=cv2.COLORMAP_VIRIDIS):

grad_model = Model(inputs=input_model.input, outputs=[input_model.get_layer(layer_name).output, input_model.output])

with tf.GradientTape() as tape:

conv_outputs, predictions = grad_model(img_array)

loss = predictions[:, class_index]

grads = tape.gradient(loss, conv_outputs)

pooled_grads = tf.reduce_mean(grads, axis=(0, 1, 2))

conv_outputs = conv_outputs[0]

heatmap = tf.reduce_mean(tf.multiply(pooled_grads, conv_outputs), axis=-1)

heatmap = tf.maximum(heatmap, 0) # Use TensorFlow's maximum function here

# Convert the tensor to a NumPy array using a TensorFlow session

with tf.compat.v1.Session() as sess:

heatmap_array = heatmap.eval()

heatmap_array /= np.max(heatmap_array)

heatmap_array = cv2.resize(heatmap_array, (img_size, img_size))

heatmap_array = np.uint8(255 * heatmap_array)

heatmap_array = cv2.applyColorMap(heatmap_array, colormap)

superimposed_img = cv2.addWeighted(cv2.cvtColor(img_array[0], cv2.COLOR_RGB2BGR), 0.6, heatmap_array, 0.4, 0)

return superimposed_img

import tensorflow as tf

import cv2

import numpy as np

import matplotlib.pyplot as plt

from tensorflow.keras.preprocessing.image import ImageDataGenerator

from tensorflow.keras.applications import VGG16

from tensorflow.keras.models import Sequential, Model

from tensorflow.keras.layers import Dense, Dropout, Flatten

from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint, LearningRateScheduler

from sklearn.metrics import classification_report, confusion_matrix

import seaborn as sns

from lime import lime_image

# ...

# Choose a random image from the test set for visualization

test_img, true_label = test_generator.next()

class_index = 0 # Choose the class index you want to visualize (corresponding to class_names)

# layer_name = 'block5_conv3' # Choose a layer from the base VGG16 model

# Use the functional API model for Grad-CAM visualization

# test_img_with_gradcam = visualize_grad_cam(model, test_img, layer_name, class_index=class_index)

# test_img_with_gradcam = visualize_grad_cam(base_model, test_img, layer_name, class_index=class_index)

layer_name = 'block5_conv2' # Choose a different layer from the base VGG16 model

test_img_with_gradcam = visualize_grad_cam(base_model, test_img, layer_name, class_index=class_index)

# Create the LIME explainer

explainer = lime_image.LimeImageExplainer()

# Display the original image, Grad-CAM visualization, and LIME explanation

fig, axes = plt.subplots(1, 3, figsize=(18, 6))

axes[0].imshow(test_img[0])

axes[0].set_title('Original Image')

axes[0].axis('off')

axes[1].imshow(test_img_with_gradcam)

axes[1].set_title(f'Image with Grad-CAM (Class {class_index})')

axes[1].axis('off')

axes[2].imshow(explanation_img)

axes[2].set_title('LIME Explanation')

axes[2].axis('off')

plt.show()

this is the error I get 

--------------------------------------------------------------------------- FailedPreconditionError Traceback (most recent call last) File ~/anaconda3/envs/flaskenv/lib/python3.9/site-packages/tensorflow/python/client/session.py:1379, in BaseSession._do_call(self, fn, *args) 1378 try: -> 1379 return fn(*args) 1380 except errors.OpError as e: File ~/anaconda3/envs/flaskenv/lib/python3.9/site-packages/tensorflow/python/client/session.py:1362, in BaseSession._do_run.._run_fn(feed_dict, fetch_list, target_list, options, run_metadata) 1361 self._extend_graph() -> 1362 return self._call_tf_sessionrun(options, feed_dict, fetch_list, 1363 target_list, run_metadata) File ~/anaconda3/envs/flaskenv/lib/python3.9/site-packages/tensorflow/python/client/session.py:1455, in BaseSession._call_tf_sessionrun(self, options, feed_dict, fetch_list, target_list, run_metadata) 1453 def _call_tf_sessionrun(self, options, feed_dict, fetch_list, target_list, 1454 run_metadata): -> 1455 return tf_session.TF_SessionRun_wrapper(self._session, options, feed_dict, 1456 fetch_list, target_list, 1457 run_metadata) FailedPreconditionError: 2 root error(s) found. (0) FAILED_PRECONDITION: Could not find variable block5_conv1_2/kernel. This could mean that the variable has been deleted. In TF1, it can also mean the variable is uninitialized. Debug info: container=localhost, status error message=Container localhost does not exist. (Could not find resource: localhost/block5_conv1_2/kernel) [[{{node model_11/block5_conv1/Conv2D/ReadVariableOp}}]] [[model_11/block5_conv3/Conv2D/ReadVariableOp/_28]] (1) FAILED_PRECONDITION: Could not find variable block5_conv1_2/kernel. This could mean that the variable has been deleted. In TF1, it can also mean the variable is uninitialized. Debug info: container=localhost, status error message=Container localhost does not exist. (Could not find resource: localhost/block5_conv1_2/kernel) [[{{node model_11/block5_conv1/Conv2D/ReadVariableOp}}]]

...

File "/Users/senu2k/anaconda3/envs/flaskenv/lib/python3.9/site-packages/tensorflow/python/framework/op_def_library.py", line 795, in _apply_op_helper op = g._create_op_internal(op_type_name, inputs, dtypes=None, File "/Users/senu2k/anaconda3/envs/flaskenv/lib/python3.9/site-packages/tensorflow/python/framework/ops.py", line 3381, in _create_op_internal ret = Operation.from_node_def(