maryzhang/image-dataset-guideline

Dataset from Google Sheets

This dataset was automatically extracted from Google Sheets with image columns.

Dataset Structure

Original Columns:

• —#
• —Text Entry
• —Topic
• —Source
• —Good/Bad Practice?
• —Reasoning Type
• —Feature Type
• —Source Link (Interview/Website/etc)
• —Manufacturing Type
• —Pontential Prompt Type
• —Explanation

Image Columns (processed):

• —Topic_local: Local path to downloaded image (originally from 'Topic')

Statistics:

• —Total records: 382
• —Image columns found: 1
• —Total images downloaded: 0

Usage

from datasets import load_dataset, Image

# Load from Hugging Face Hub
dataset = load_dataset("your-username/your-dataset-name")

# Or load locally
from datasets import Dataset
import json

with open("dataset.json", "r") as f:
    data = json.load(f)

dataset = Dataset.from_list(data)

# Cast image columns
image_columns = ['"Topic_local"']
for col in image_columns:
    if col in dataset.column_names:
        dataset = dataset.cast_column(col, Image())

Image Processing Notes

Images were automatically downloaded from the following URL columns:

• —Topic

Source

Extracted from Google Sheets on 2025-11-30 Sheet URL: Google Sheets