Dataset Description -
This dataset contains a synthetic multimodal collection of product images and corresponding advertising texts.
Each sample includes a product image and structured metadata such as product name, category, tagline, ad copy, and target audience.
Data Generation -
The dataset was generated using pre-trained Hugging Face models.
Images were created using Stable Diffusion, while advertising texts were generated using the Phi-2 language model.
EDA -
The dataset contains 120… See the full description on the dataset page:
https://huggingface.co/datasets/Barvero/adautogen-multimodal.
