datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
perfume-description
Perfume Description Dataset
Dataset Description
The Perfume Description Dataset contains perfume descriptions generated by LLMs using different prompt engineering strategies, annotated by humans for preference. The dataset was created to analyze which generation methods produce the most preferred descriptions and whether ML models can predict human preferences for creative product descriptions.
Dataset Summary
Total Perfumes: 255
Total Annotations (T1): 664… See the full description on the dataset page: https://huggingface.co/datasets/MrBob23/perfume-description.DE-PERFUME
🌸 DE-PERFUME
Multimodal Synthetic Perfume Dataset (Text + Image)
✨ Overview
DE-PERFUME is a synthetic multimodal dataset of perfume concepts.Each sample links a natural-language user request to structured attributes and an AI-generated perfume bottle image.
📦 Records: 100
🧠 Modalities: Text + Image
🎯 Primary use case: Text → Image retrieval
📂 Dataset Contents
perfume_dataset.csv — main metadata table
perfume_images.zip — all generated… See the full description on the dataset page: https://huggingface.co/datasets/Omrynadiv/DE-PERFUME.imnet1k_perfume_essencePerfume