The DL.ipynb file uploaded on moodle uses a pre-trained mobilenet model to run inference on flowers dataset using Pyspark.
1. Modify the above code to run inference on CIFAR 10 dataset using Pyspark. (7)
2. Try out a few different models pre-trained on Imagenet and report which one works better (calculating exact accuracy is difficult as the class names in imagenet and CIFAR
10 dataset don’t exactly match, but still printing out the predictions for a few points and
looking at the class names should give a hint). (3)
Zip the code file along with a pdf having the observations about different models.
