py -m scripts.retrain --bottleneck_dir=tf_files/bottlenecks --how_many_training_steps=500 --model_dir=tf_files/models/ --summaries_dir=tf_files/training_summaries/"mobilenet_0.50_224" --output_graph=tf_files/retrained_graph.pb --output_labels=tf_files/retrained_labels.txt --architecture="mobilenet_0.50_224" --image_dir=tf_files/products