Welcome to the demo page accompanying the Gold-Caps dataset repository.
This page demonstrates the use of the "gpt-4-1106-preview" OpenAI API endpoint for generating captions with a prompt solely referencing the track title and artist name. The webpage has been tested with Google Chrome
💬 General description: Give a general description of the track <title> by artist_name> in one sentence. Don't mention the title or artist.
☕️ Vibe: Describe the vibe of the track <title> by artist_name> in one sentence. Don't mention the title or artist.
🎼 Arrangement: Describe the arrangement of the track <title> by artist_name> in one sentence. Don't mention the title or artist.
🎛️ Production: Describe the production style of the track <title> by artist_name> in one sentence. Don't mention the title or artist.
🤖 AI instruction: What prompt would you give an AI text-to-music generation if you wanted the output to sound like the track <title> by artist_name>. One sentence only. Don't mention the title or artist.
🎬 Movie scene: Describe an imagined movie scene that would be set to the track <title> by artist_name> in one sentence. Don't mention the title or artist.
Short audio previews are provided for research purposes.