Datasetpublic

TempoFunk/small

TempoFunk Small 7.8k samples of metadata and encoded latents & prompts of random videos. Data format Video frame latents Numpy arrays 120 frames, 512x512 source size Encoded shape (120, 4, 64, 64) CLIP (openai) encoded prompts Video description (as seen in metadata) Encoded shape (77,768) Video metadata as JSON (description, tags, categories, source URL, etc.)

sourceHugging Faceagpl-3.0updated 3y agoView on Hugging Face

9likes69kdownloads

rootREADME.md

README.md26 linesDownload Raw Back to root

1---2task_categories:3- text-to-video4- text-to-image5- video-classification6- image-classification7language:8- en9size_categories:10- 1K<n<10K11license: agpl-3.012---13 14# TempoFunk Small15 167.8k samples of metadata and encoded latents & prompts of random videos.17 18## Data format19- Video frame latents20  - Numpy arrays 21  - 120 frames, 512x512 source size22  - Encoded shape (120, 4, 64, 64)23- CLIP (openai) encoded prompts24  - Video description (as seen in metadata)25  - Encoded shape (77,768)26- Video metadata as JSON (description, tags, categories, source URL, etc.)