Training & validation accuracy increasing & training loss is decreasing - Validation Loss is NaN

Question

I am training a classifier model on cats vs dogs data. The model is a minor variant of ResNet18 & returns a softmax probability for classes. However, I am noticing that the validation loss is majorly NaN whereas training loss is steadily decreasing & behaves as expected. Training & Validation accuracy increase epoch by epoch.

Epoch 1/15
312/312 [==============================] - 1372s 4s/step - loss: 0.7849 - accuracy: 0.5131 - val_loss: nan - val_accuracy: 0.5343
Epoch 2/15
312/312 [==============================] - 1372s 4s/step - loss: 0.6966 - accuracy: 0.5539 - val_loss: 13989871201999266517090304.0000 - val_accuracy: 0.5619
Epoch 3/15
312/312 [==============================] - 1373s 4s/step - loss: 0.6570 - accuracy: 0.6077 - val_loss: 747123703808.0000 - val_accuracy: 0.5679
Epoch 4/15
312/312 [==============================] - 1372s 4s/step - loss: 0.6180 - accuracy: 0.6483 - val_loss: nan - val_accuracy: 0.6747
Epoch 5/15
312/312 [==============================] - 1373s 4s/step - loss: 0.5838 - accuracy: 0.6852 - val_loss: nan - val_accuracy: 0.6240
Epoch 6/15
312/312 [==============================] - 1372s 4s/step - loss: 0.5338 - accuracy: 0.7301 - val_loss: 31236203781405710523301888.0000 - val_accuracy: 0.7590
Epoch 7/15
312/312 [==============================] - 1373s 4s/step - loss: 0.4872 - accuracy: 0.7646 - val_loss: 52170.8672 - val_accuracy: 0.7378
Epoch 8/15
312/312 [==============================] - 1372s 4s/step - loss: 0.4385 - accuracy: 0.7928 - val_loss: 2130819335420217655296.0000 - val_accuracy: 0.8101
Epoch 9/15
312/312 [==============================] - 1373s 4s/step - loss: 0.3966 - accuracy: 0.8206 - val_loss: 116842888.0000 - val_accuracy: 0.7857
Epoch 10/15
312/312 [==============================] - 1372s 4s/step - loss: 0.3643 - accuracy: 0.8391 - val_loss: nan - val_accuracy: 0.8199
Epoch 11/15
312/312 [==============================] - 1373s 4s/step - loss: 0.3285 - accuracy: 0.8557 - val_loss: 788904.2500 - val_accuracy: 0.8438
Epoch 12/15
312/312 [==============================] - 1372s 4s/step - loss: 0.3029 - accuracy: 0.8670 - val_loss: nan - val_accuracy: 0.8245
Epoch 13/15
312/312 [==============================] - 1373s 4s/step - loss: 0.2857 - accuracy: 0.8781 - val_loss: 121907.8594 - val_accuracy: 0.8444
Epoch 14/15
312/312 [==============================] - 1373s 4s/step - loss: 0.2585 - accuracy: 0.8891 - val_loss: nan - val_accuracy: 0.8674
Epoch 15/15
312/312 [==============================] - 1374s 4s/step - loss: 0.2430 - accuracy: 0.8965 - val_loss: 822.7968 - val_accuracy: 0.8776

I checked for the following -

• Infinity/NaN in validation data
• Infinity/NaN caused when normalizing data (using tf.keras.applications.resnet.preprocess_input)
• If the model is predicting only one class & hence causing loss function to behave oddly

Training code for reference -

optimizer = tf.keras.optimizers.Adam(learning_rate=5e-3)
model = Resnet18(NUM_CLASSES=NUM_CLASSES) # variant of original model
model.compile(optimizer=optimizer, loss="categorical_crossentropy", metrics=["accuracy"])

history = model.fit(
    train_dataset,
    steps_per_epoch=len(X_train) // BATCH_SIZE,
    epochs=EPOCHS,
    validation_data=valid_dataset,
    validation_steps=len(X_valid) // BATCH_SIZE,
    verbose=1,
)

The most relevant answer I found was the last paragraph of the accepted answer here. However, that doesn't seem to be the case here as validation loss diverges by order of magnitudes compared to training loss & returns nan. Seems like the loss function is misbehaving.