README.md

metadata

annotations_creators:
  - machine-generated
language_creators:
  - machine-generated
language:
  - en
license:
  - mit
multilinguality:
  - monolingual
size_categories:
  - n<100
task_categories:
  - text-generation
task_ids:
  - language-modeling

Streamlit Dashboard Dataset

80 Streamlit dashboard Python scripts for training code generation models.

Structure

• data/train-00000-of-00001.parquet — Main dataset in Parquet format
• README.md — Dataset card

Columns

Column	Type	Description
filename	string	File name of the dashboard
title	string	Human-readable title
code	string	Full Python source code
libraries	string	Semi-colon separated libraries used
chart_types	string	Semi-colon separated chart types
metric_count	int32	Number of st.metric() calls
has_dataframe	bool	Whether st.dataframe() is used
file_size	int32	File size in bytes
token_count	int32	Approximate token count

Usage

from datasets import load_dataset

dataset = load_dataset("sanjaymalladi/streamlit-dataset", split="train")
print(dataset[0]["filename"])
print(dataset[0]["code"][:200])

Domains

Sales, Marketing, Crypto, Health, Weather, Ecommerce, Social Media, Sports, Education, Real Estate, Music, Gaming, Travel, Food, Environment, HR, Projects, Support, Supply Chain, Energy, Transport, Public Health, Demographics, Lab, SEO, Email, CRM, Inventory, Manufacturing, Logistics, Agriculture, Climate, Ocean, Wildlife, Space, Film, Publishing, Podcast, YouTube, Esports, Fitness, Sleep, Mindfulness, NFT, AI, Security, News, University, Restaurant, Astronomy, Stocks, VC, Retail, Pharma, IoT, COVID, Elections, ML Models

Chart Libraries

• Plotly — 75 files (scatter, bar, line, pie, funnel, heatmap, waterfall, polar, etc.)
• Altair — 3 files (declarative statistical charts)
• Streamlit built-in — 2 files (st.line_chart, st.bar_chart, st.area_chart)