Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Held-out Prompts

Benchmarks

Task NameDataset NameSOTA ResultTrend
Text-to-Music configuration generationHeld-out 200 prompts (test)
CLAP Score (mean)0.163
6
Story Generation Quality180 held-out prompts (ID 1–4k)
Story Quality Score57.4
1
Showing 2 of 2 rows