sunbird_stable_difussion_photos Dataset

This dataset contains images with corresponding text captions, prepared for Stable Diffusion fine-tuning.

Dataset Details

• Total Images: 295
• Image Size: 512x512
• Caption Strategy: ai
• AI Model Used: nlpconnect/vit-gpt2-image-captioning

Dataset Structure

{
  "image": <PIL.Image>,
  "text": <caption_string>,
  "file_name": <original_filename>
}

Usage

Loading the Dataset

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("yigagilbert/sunbird_stable_difussion_photos")

# Access images and captions
for example in dataset["train"]:
    image = example["image"]
    caption = example["text"]
    filename = example["file_name"]

For Stable Diffusion Training

This dataset is ready for use with popular Stable Diffusion training scripts:

• Dreambooth: Use the image-caption pairs directly
• LoRA Training: Compatible with most LoRA training scripts
• Textual Inversion: Can be used for concept learning

Citation

If you use this dataset, please cite:

@dataset{yigagilbert_sunbird_stable_difussion_photos,
  title={sunbird_stable_difussion_photos Dataset},
  author={Your Name},
  year={2025},
  url={https://huggingface.co/datasets/yigagilbert/sunbird_stable_difussion_photos}
}

License

This dataset is released under the MIT License.