Datasets:
| license: other | |
| task_categories: | |
| - text-to-video | |
| language: | |
| - en | |
| tags: | |
| - video | |
| - webdataset | |
| - parquet | |
| private_backup: true | |
| # Connan 30K Video Caption Dataset | |
| Private Hugging Face backup of 30,617 short anime video clips and English visual captions. | |
| ## Layout | |
| ```text | |
| metadata/manifest.parquet # sample index, captions, original paths, shard/member mapping | |
| wds/*.tar # WebDataset shards, about 10GB each | |
| dataset_info.json # summary metadata | |
| ``` | |
| Each WebDataset sample uses a stable key such as `001_shot_00003` and contains: | |
| ```text | |
| 001_shot_00003.mp4 | |
| 001_shot_00003.json | |
| ``` | |
| The JSON sidecar contains the caption and original source paths. The Parquet manifest is the canonical table for captions and shard lookup. | |
| ## Summary | |
| - Samples: 30617 | |
| - Shards: 10 | |
| - Caption language: English | |
| - Recommended LoRA trigger: `DCANIME` | |
| Captions are generic visual descriptions and intentionally avoid character names, franchise names, and trigger words. | |