Dataset Viewer

About

Dataset Viewer is an MCP server designed to interact with the Hugging Face Dataset Viewer API, granting users the ability to browse, filter, and analyze datasets hosted on the Hugging Face Hub. It uses the `dataset://` URI scheme for dataset access, supports configurations and splits, offers paginated access, handles authentication for private datasets, and provides tools for searching, filtering (using SQL-like conditions), and statistical analysis. The server also allows users to download entire datasets in Parquet format, making it a comprehensive solution for working with Hugging Face datasets.

Key Features

• Browse Hugging Face datasets with configurations and splits.
• Filter dataset contents using SQL-like WHERE clauses.
• Search for specific text within datasets.
• Retrieve dataset statistics for analysis.
• Download entire datasets in Parquet format.

Use Cases

• Validating the existence and accessibility of a dataset on the Hugging Face Hub.
• Searching for specific examples within a dataset to understand its content.
• Filtering a dataset to create a subset for training a machine learning model.