LAION releases CLAP2 dataset with 1M+ hours and 365M audio clips
lhoestq · x · 2026-11-12
LAION has released laion-audio-xl, the CLAP2 dataset, on Hugging Face: over 1 million hours and 365 million audio clips curated from YouTube, Common Crawl, web collections, and 300+ academic datasets.
- Organized into open, sharealike, and noncommercial license configs, with per-clip licenses and credits.
- The links config contains only URLs, timestamps, and machine-made annotations — no audio; users must comply with the underlying media's terms when fetching content.
Related event: LAION Releases Web-Scale Audio-XL Dataset and CLAP2 Models(3 posts)→
More from Research
- ICLR abstract word-frequency shift: 'novel' and 'comprehensive' plummet as LLM fingerprints fade — y_m_asano · 2026-11-12
- Evolving Infra: agent-adaptive LLM serving cuts latency up to 4.1x vs NVIDIA Dynamo — mit-han-lab · 2026-11-12
- Isomorphic Labs President on Using AI to Turn Drug Discovery into Precision Engineering — No Priors · 2026-11-12
- CreativityPrism benchmark scores 17 frontier and open LLMs across 8 creative tasks — DanielKhashabi · 2026-11-12
- KQuant: open-source MLX quants cut Qwen KL divergence up to 3.7x vs public 4-bit — m4rkGS · 2026-11-12
- Possible eval leak: all frontier models fail a GPQA Diamond question whose official answer seems wrong — qberthet · 2026-11-12