Dataset · Hugging Face

JinchuanTian/mmau_generation_test

Name: mmau_generation_test
Creator: JinchuanTian
License: other
Keywords: task_categories:text-to-audio, language:en, license:other, size_categories:n<1K, format:audiofolder, modality:audio, library:datasets, library:mlcroissant, region:us, text-to-audio, speechlm, sft, audio-generation

MMAU-Mini Text-to-Audio SFT (realistic, v1) Supervised fine-tuning data for open-ended text-to-audio generation. Each example pairs a natural-language user request with an assistant turn that reasons about the request and then produces the target audio. Built by reverse-engineering plausible user requests from the rich captions of the MMAU-Mini test set. 995 examples across three categories: music (331), sound (331), speech (333). Realistic prompts only (requests that directly… See the full description on the dataset page: https://huggingface.co/datasets/JinchuanTian/mmau_generation_test.

View on Hugging FaceReport an issue

Author / organization	JinchuanTian
License (as declared)	other
Downloads	126 (as reported by Hugging Face)
Likes	0
Created	31 May 2026
Last updated at source	31 May 2026

text-to-audiospeechlmsftaudio-generation

Embed this entry

<iframe src="https://silytics.si/embed/item/b2a0d960-8126-46b0-b3fe-4235ae9a4ba9" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 2 days

126 now (-7 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

More from JinchuanTian

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.