Dataset · svjack
Pokemon images with English and Chinese captions dataset
svjack/pokemon-blip-captions-en-zh
Small set of Pokemon images with bilingual captions for training text-to-image models.
What's inside
Contains fewer than a thousand Pokemon images from the FastGAN set, supplemented with text descriptions. English captions were generated by a pretrained BLIP model and then translated into Chinese using DeepL. The data is packed into a parquet format file.
What you can do with it
- Training LoRA adapters for Pokemon-style image generation
- Testing diffusion model fine-tuning pipelines on small bilingual samples
- Tasks
- Text → image
- Languages
- English, Chinese
- Size
- up to 1K rows
- Downloads this month
- 980
- Likes
- 54
Models trained on it
How to load it
from datasets import load_dataset
ds = load_dataset("svjack/pokemon-blip-captions-en-zh")