HW9 - Deep Learning with PyTorch (50 points)

Due Sunday 06/05 at 11:59 pm.

You may submit any part of the assignment as many times as you want before the cutoff. Everything must be submitted on Gradescope. Please make sure you are familiar with the collaboration policy on the syllabus.

Note that as mentioned on Ed there will be no late days allowed for this assignment. Therefore, the hard deadlines for the homework are:

  • Written and Quiz: Sunday 11:59pm (6/5)
  • Programming: Thursday 11:59pm (6/9)

These deadlines are non-negotiable to allow feedback to be given back in time for the final. If you have specific circumstances that make it difficult to submit by these dates, please reach out to Pemi.

Assignment Structure

For each homework assignment, there will usually be three things to submit:

  • A Conceptual portion that asks you to solve conceptual questions about that week’s materials

    • Quiz: You must take it on Gradescope. You have unlimited attempts. The quiz will be autograded but results won’t be released until after the deadline.
    • Written questions: You must submit a typeset PDF on Gradescope and attach the correct pages on Gradescope.
  • A Programming portion that asks you to answer questions or do an analysis involving programming. This counts towards your Programming portion of your final grade. You can write the code locally on EdStem and verify the correctness with an autograder, but you must submit the code with a .py file on Gradescope to finalize your grade.



Please find the written assignment questions here: hw9.pdf

Submit the written questions directly on Gradescope: Written submission


Submit the quiz directly on Gradescope: Quiz submission


Find the Jupyter notebook of the programming assignment here: Jupyter notebook

Note that for this programming assignment there is no autograder. Once you finish the assignment, download the notebook from Colab as a .ipynb notebook (File > Download > Download .ipyb) and upload it on Gradescope. Your assignment should be named cse416_hw9.ipynb.

Additionally, understanding PyTorch takes time. For this homework, we’ve made PyTorch coding as straightforward as possible. You only need to define different neural network architectures. Make sure to read Pytorch documentations to understand how to do so. Furthermore, be aware that training neural networks is time-consuming. Allocate proper time to do so.