Ploba logo

Find the right AI tool for the job: apps, agents, APIs, MCP servers and datasets, explained plainly.

Get the AI Digest

Discover

  • Explore
  • Top Charts
  • Categories
  • Agents
  • MCP servers
  • Developer tools

Learn & community

  • Learn AI
  • AI Glossary
  • Blog
  • Hackathons
  • People
  • AI Digest

For makers

  • List your tool
  • Docs
  • Claim a listing
  • Advertise
  • Write for us

Company

  • About
  • Careers
  • Changelog
  • Trust
  • Security
  • Contact

© 2026 Ploba. All rights reserved.

  • Privacy
  • Terms
  • Cookies
  • Advertising
  • All policies
llms.txt
Ploba logo
DiscoverHackathonsPeopleLearnAI Digest
Sign inGet started
HomeTop ChartsExploreCategories
TypesWeb apps1,894Agents585Plugins416APIs376Mobile apps152Datasets85
AdvertisingYour product hereReach people browsing AI tools on Ploba.Advertise hereAdvertisingYour product hereReach people browsing AI tools on Ploba.Advertise here
  1. Explore
  2. /
  3. Data
  4. /
  5. WildSongBench
Data
Wi

WildSongBench

by M-A-P

WildSongBench is a benchmark of 192 song-generation prompts, 94 in Chinese and 98 in English, from the M-A-P group. It ships prompts, exact inputs, seeds, reference scores and a small evaluator, but no audio.

Dataset·Free

View on Hugging Face(opens in a new tab)

About WildSongBench

Key features of WildSongBench

  • 192 song-generation prompts: 94 Chinese and 98 English
  • Style, lyrics and exact seeds provided for every prompt
  • Minimal evaluator for SongBench, Q3O, MuQ-MuLan, AllMusicCaps and phoneme error rate
  • Recorded reference scores for 17 settings
  • Step-by-step reproduction instructions for the standard YuE2 run
  • Loads with the Hugging Face datasets library

Who WildSongBench is for

  • Researchers evaluating full-song music generation models
  • Teams comparing open-weights and proprietary song generators
  • Developers reproducing YuE2 benchmark results

How to use WildSongBench

  1. 1

    Step 1: Load the test split with load_dataset("m-a-p/WildSongBench", split="test").

  2. 2

    Step 2: Read each record's prompt and the style, lyrics and seeds for generation.

  3. 3

    Step 3: Generate songs with your model for all 192 prompts, since audio is not included.

  4. 4

    Step 4: Install the evaluator from the repository and run the prepare step.

  5. 5

    Step 5: Run the score commands and compare against the reference scores in the card.

WildSongBench FAQ

Does WildSongBench include audio?

No. It provides prompts, exact inputs and seeds, reference scores and an evaluator. You generate the audio yourself.

How many prompts are there and in which languages?

There are 192 prompts: 94 in Chinese and 98 in English.

Which metrics does the evaluator cover?

SongBench, Q3O, MuQ-MuLan, AllMusicCaps and phoneme error rate.

What hardware is needed to run the evaluation?

Q3O requires an 80 GB GPU, and the other metrics support 24 GB. The card also lists Linux and a fresh Python environment.

Details

Pricing
Free
Type
Dataset
Categories
DataAudioResearch
Added
Sep 17, 2026
M-PublisherM-A-P

Built from public sources

We put this page together from WildSongBench’s own website and docs so you can compare it with other tools. Ploba isn’t affiliated with WildSongBench.

Work at WildSongBench? Claim the listing to manage it yourself.

Something wrong here?

Related tools

See more
PaParsVoiceMohammad J. Ranjbar · FreeParsVoice is a Persian speech and text corpus for training multi-speaker text-to-speech systems. It holds about 1.36 million audio segments totaling roughly 2,178 hours.Data · FreeFLEURS iconFLEURSGoogle · FreeFLEURS is Google's multilingual speech dataset covering 102 languages, built as the speech version of the FLoRes machine translation benchmark. It supports speech recognition, language identification and speech-text retrieval.Data · FreeMatbench Discovery iconMatbench DiscoveryMaterials Project · FreeMatbench Discovery is a leaderboard that benchmarks machine-learning force fields on crystal discovery, geometry optimization, phonons, molecular dynamics and diatomics.Data · Free