mir-dataset-loaders/mirdata

Python library for working with Music Information Retrieval datasets

What it solves

This project addresses the fragmentation and inconsistency in how Music Information Retrieval (MIR) datasets are accessed and processed. It provides a unified interface to download, validate, and load various audio datasets, ensuring that researchers can work with them in a consistent, reproducible manner.

How it works

It acts as a standardized wrapper for numerous MIR datasets. Users initialize a specific dataset loader, which then handles the downloading of files, verifies that the dataset is complete and intact, and parses annotations into a common format that is compatible with standard evaluation tools like mir_eval.

Who it’s for

This library is designed for researchers and developers working in audio machine learning and music information retrieval who need to streamline their data pipeline and ensure reproducibility across different experiments.

Highlights

  • Standardized API for downloading and validating diverse audio datasets.
  • Consistent annotation parsing compatible with mir_eval.
  • Support for a wide range of popular datasets like MAESTRO, Guitarset, and AcousticBrainz.
  • Facilitates reproducible research by providing track-level metadata and consistent data structures.

Related

  • Project
    timescale/pg_vectorsearch

    A PostgreSQL extension providing a high-performance, scalable vector search index (PRISM) that uses hierarchical clustering and quantization to outperform standard HNSW and IVFFlat methods.

  • Project
    grobidOrg/grobid-client-python

    A Python library and CLI that talks to a GROBID server to extract structured metadata, references, full‑text, figures, tables, etc., from PDFs and patents. It supports parallel processing, archive/S3 streaming, and can output TEI XML, CORD‑19‑style JSON, or Markdown.

  • Project
    qdrant/rust-client

    A native Rust client for Qdrant, enabling high-performance vector search and database management within Rust applications.

  • Project
    runekaagaard/mcp-alchemy

    MCP Alchemy is an MCP server that connects Claude Desktop to various SQL databases, enabling the AI to explore schemas, write queries, and perform data analysis directly.