HW0

• Release data: 01/07/2025
• Due date: 01/21/2025 at 11:59 PM.

HW0 consists of two parts:

• Part 1: Search and Value Iteration (50pts)
• Part 2: LM training with Huggingface (50pts)

You can find the starter code and instruction in the course’s GitHub repository: https://github.com/pearls-lab/ai-agents-course

Submission

Submissions should be done on Gradescope. You need to submit code for Part 1 and Part 2, and a PDF file for Part 2.

Submit your code

Please submit the following 4 files to the assignment “HW 0 - Code” on Gradescope:

• search.py
• searchAgents.py
• valueIterationAgents.py
• hw0-part-2.ipynb

Do not zip or compress your files. Submit them as individual files. Part 1 will be automatically graded by the Gradescope system. The score displayed immediately after submission will be your final score for Part 1.

Submit your PDF (only for Part 2)

1. Re-run all cells in order. Ensure that all cell outputs are displayed correctly.
2. Convert the notebook to PDF:
   • Download the .ipynb file to your local machine.
   • Use a tool like nbconvert to convert the notebook to PDF.
   • Alternatively, if you encounter issues with nbconvert, you can save the notebook webpage as a PDF.
3. Review the PDF file: Look at the PDF file and make sure all your codes are displayed correctly.
4. Submit your PDF and the notebook file on Gradescope: “HW 0 - Part 2 PDF”