Skip to content

TUTORIAL.md

Latest commit

 

History

History
132 lines (86 loc) · 4.66 KB

TUTORIAL.md

File metadata and controls

132 lines (86 loc) · 4.66 KB
Raw

Memory Garden Tutorial

You'll be turning your personal memories into AI-generated flower art — all in the browser.

Step 0 — Download and open the project

  1. Go to the GitHub repository link provided by your TA.
  2. Click the green Code button, then click Download ZIP.
  3. Find the downloaded .zip file (usually in your Downloads folder) and unzip it.
  4. Open Cursor (our code editor for this workshop).
  5. In Cursor, go to File → Open Folder and select the unzipped project folder.

You should now see the project files (like server.py, TUTORIAL.md, etc.) in Cursor's sidebar.

Step 1 — Install uv (one time only)

Mac (Terminal):

curl -LsSf https://astral.sh/uv/install.sh | sh

Windows (PowerShell):

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

After installing, restart your terminal (or run source ~/.zshrc on Mac).

Verify it works:

uv --version

Step 2 — Set up your API key

  1. In Cursor's sidebar, find the file called .env.example and right-click it, then select Rename.
  2. Rename it to .env (just remove the .example part).
  3. Open the .env file, replace your_api_key_here with the API key provided by your TA.
  4. Save the file (Ctrl+S on Windows / Cmd+S on Mac).

Step 3 — Start the Memory Garden

Open your terminal in cursor, navigate to the project folder, and run:

uv run --with flask --with replicate server.py

Your browser will automatically open http://localhost:8080. You should see a flower canvas on the left and a control panel on the right.

If the browser doesn't open automatically, go to http://localhost:8080 manually.

Mac users: if you get a 403 error, macOS AirPlay may be using port 5000. Our server uses port 8080 to avoid this.

Step 4 — Load your memories

  1. In the Student Memory dropdown at the top-right, find and select your name.
  2. Click Load Student Memories.
  3. Wait 10–30 seconds — the system uses DeepSeek V3.1 (an AI model) to analyze your memories and assign emotional features to each one.
  4. Once loaded, your first memory flower will appear on the canvas.

The Emotional Features panel shows the five dimensions that shape your flower:

Feature What it means How it affects the flower
Valence Sad (-1) to Happy (+1) Color warmth, head droop
Impact Calm (0) to Intense (1) Size, petal shape, outline weight
Togetherness Solo (0) to Social (1) Number of petals, ring spacing
Motion Still (0) to Lively (1) Sway speed, hatching texture
Length Brief (0) to Detailed (1) Stem height, petal curl

Step 5 — Explore your flowers

Press R on your keyboard (or refresh) to shuffle through your memories. Each memory produces a unique flower — the shape, color, and texture all come from the emotional analysis of that memory.

Take a moment to see how different memories create different flowers. A sad, intense memory will look very different from a happy, calm one.

Step 6 — Generate AI art

Once you find a flower you like:

  1. Choose Flower Head Only or With Stem in the export mode dropdown.
  2. Type a prompt describing the style you want. You must include "photorealistic" so we have a general theme.
  3. Click Generate with Nano Banana.
  4. Wait 10–30 seconds for the AI to transform your flower.
  5. The result will appear below the button.

Example prompt:

Example prompt: “Turn this into a photorealistic exotic flower. Make it look like it grew in a lush tropical rainforest”

Example prompt: “Make this into a photorealistic arctic flower.”

Feel free to be creative with your prompt — try different styles and see what happens but always remember a consistent prompt as part of the whole prompt: "Make this into a photorealistic".

Your generated images are automatically saved to the MemoryGarden/ folder. Take sometime to make 3-5 images and look at them carefully.

Step 7 — Submit

  1. Find your 3-5 generated images in the MemoryGarden/ folder.
  2. Rename each file with your full name and a number — for example: John_Smith.png
  3. Upload 1 image you like to the shared Google Drive folder: https://drive.google.com/drive/folders/1wCa1v2l9Q7qvPiXCMJVWwtCFwCMVwOcZ

Quick reference

Action How
Start the server uv run --with flask --with replicate server.py
Shuffle memories Press R
Generate AI art Type prompt + click Generate with Nano Banana
Stop the server Press Ctrl+C in the terminal