NLPDL - Fall 2025 Homework

Welcome to the official repository for the NLP course for Fall 2025. This repository contains all the homework assignments for the course.

Repository Structure

This repository is organized into directories, with each directory corresponding to a specific homework assignment.

NLPDL-2025Fall/
├── .gitignore
├── README.md
├── hw0_hello_world/
│   ├── __init__.py
│   ├── pipeline.py
│   ├── README.md
│   ├── test_pipeline.py
│   └── pyproject.toml      # Assignment-specific environment (uv)
└── hw1_bpe_and_lm/
    ├── assignment1.md
    ├── pyproject.toml      # Assignment-specific environment (uv)
    └── ...

• hw0_hello_world/: An introductory assignment to familiarize you with the homework submission and testing process.

• hw1_.../: The first main assignment (to be added).

• ...: Subsequent assignments.

Each homework directory is a self-contained Python package and includes:

• A README.md with specific instructions for that assignment.

• One or more Python files (e.g., pipeline.py) where you will implement your code.

• A testing file (e.g., test_pipeline.py) containing pytest tests to help you verify your implementation.

Getting Started (for Students)

Prerequisites

This project uses uv for Python package management. uv is a fast Python package and project manager that provides better dependency resolution and faster installations compared to pip.

Install uv:

# On macOS and Linux
curl -LsSf https://astral.sh/uv/install.sh | sh

# On Windows
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

# Or using pip
pip install uv

Setup Instructions

1. Clone the repository:

git clone <repository_url>
cd <repo_name>

2. Set up your Python environment (per assignment): Each homework directory is its own uv project. Work inside the assignment directory so uv can manage a dedicated .venv/ for that assignment.

# Example: Homework 0
cd hw0_hello_world
uv sync   # creates .venv/ in hw0_hello_world and installs deps

# Example: Homework 1
cd ../hw1_bpe_and_lm
uv sync

This will:

• Create a virtual environment in the assignment directory’s .venv/
• Install a compatible Python version (e.g., 3.12) if needed
• Install that assignment’s dependencies

3. Complete the assignment: Navigate to the specific homework directory (e.g., cd hw0_hello_world) and follow the instructions in its README.md to complete your implementation in the provided Python files.

4. Run tests: Run tests from within each assignment directory (recommended), or use uv’s directory flag.

# From inside an assignment directory
uv run pytest

# Or from the repo root using the assignment directory
uv run --directory hw0_hello_world pytest
uv run --directory hw1_bpe_and_lm pytest

Note: Use uv run within the assignment’s directory (or --directory) to ensure the correct per-assignment environment is used.

Additional uv Commands

# Add a new dependency
uv add <package_name>

# Remove a dependency  
uv remove <package_name>

# Run any Python command in the assignment environment
uv run python <script.py>

# Activate the virtual environment manually (optional, from assignment dir)
source .venv/bin/activate  # Windows: .venv\Scripts\activate

You should see all tests passing if your implementation is correct.

Homework Submission

Important: Do NOT submit the entire repository. Submit only the specific homework directory. See SUBMISSION_GUIDELINES.md for details.

Preferred: per-assignment submission scripts

From inside each assignment directory, run:

# hw0
cd hw0_hello_world
./make_submission.sh <LASTNAME> <FIRSTNAME> <STUDENTID>

# hw1
cd ../hw1_bpe_and_lm
./make_submission.sh <LASTNAME> <FIRSTNAME> <STUDENTID>

Each script:

• Runs tests via the assignment’s uv environment
• Creates hwX_submission_LASTNAME_FIRSTNAME_STUDENTID.zip
• Prints contents for verification

Alternative unified helper (from repo root):

./submit_hw.sh <hw_directory> <LASTNAME> <FIRSTNAME> <STUDENTID>

Refer to the guidelines for exclusions and large files.