dsfetch is a lightweight, CLI-first tool designed for machine learning practitioners who want a frictionless way to search, preview, and download datasets from multiple platforms—like Kaggle, Hugging Face, and UCI—without leaving the terminal. With a clean, unified interface and support for platform-specific identifiers, dsfetch streamlines dataset discovery and retrieval, making experimentation faster, reproducible, and less dependent on the browser. Whether you’re prototyping a new model or curating datasets for training, dsfetch gets you the data you need—fast.
- 🔹 CLI interface to input Kaggle dataset handles
- 🔹 Automated dataset download using
kagglehub
- 🔹 Directory-based organization under a central
/datasets
folder - 🔹 Smart detection of file vs. directory downloads
- 🔹 Optional renaming and clean-up logic
- 🔹 Local environment ready — no cloud dependency
The long-term vision is to evolve this into a general-purpose dataset orchestration framework — integrating intelligent agents, metadata tracking, and reproducible workflows — so developers can spend less time managing datasets and more time building great models.