REPORT5_Modified_Google_ML-Crash-Course_MNIST_model.py

I modified the MLP-based model in the previously-corrected Google Google-ML-Crash-Course_MNIST_model.py  script, resolved version conflicts, and increased the Val Accuracy up to abut 98%

=================================================================
Total params: 32XXXXX (12.XX MB)
Trainable params: 32XXXXX (12.XX MB)
Non-trainable params: 0 (0.00 Byte)
_________________________________________________________________

================================================================================
STARTING MODEL TRAINING
================================================================================
Epoch 1/10
118/118 [==============================] - 69s 540ms/step - loss: 0.8746 - accuracy: 0.7229 - val_loss: 0.2880 - val_accuracy: 0.9208
Epoch 2/10
118/118 [==============================] - 64s 543ms/step - loss: 0.2855 - accuracy: 0.9123 - val_loss: 0.1863 - val_accuracy: 0.9440
Epoch 3/10
118/118 [==============================] - 62s 524ms/step - loss: 0.2041 - accuracy: 0.9368 - val_loss: 0.1417 - val_accuracy: 0.9604
Epoch 4/10
118/118 [==============================] - 60s 507ms/step - loss: 0.1575 - accuracy: 0.9508 - val_loss: 0.1218 - val_accuracy: 0.9644
Epoch 5/10
118/118 [==============================] - 59s 502ms/step - loss: 0.1189 - accuracy: 0.9653 - val_loss: 0.0925 - val_accuracy: 0.9712
Epoch 6/10
118/118 [==============================] - 61s 516ms/step - loss: 0.1041 - accuracy: 0.9696 - val_loss: 0.0888 - val_accuracy: 0.9716
Epoch 7/10
118/118 [==============================] - 62s 531ms/step - loss: 0.0876 - accuracy: 0.9740 - val_loss: 0.0887 - val_accuracy: 0.9724
Epoch 8/10
118/118 [==============================] - 62s 523ms/step - loss: 0.0744 - accuracy: 0.9775 - val_loss: 0.0786 - val_accuracy: 0.9768
Epoch 9/10
118/118 [==============================] - 68s 572ms/step - loss: 0.0693 - accuracy: 0.9775 - val_loss: 0.0765 - val_accuracy: 0.9796
Epoch 10/10
118/118 [==============================] - ETA: 0s - loss: 0.0568 - accuracy: 0.9821
---------------------------------------------------------------------------