Datasets:

Base-to-Novel: ImageNet-1K, Caltech101, Oxford Pets, StanfordCars, Flowers102, Food101, FGVC Aircraft, SUN397, DTD, EuroSAT, UCF101.

Domain Generalization: ImageNet-V2, ImageNet-Sketch, ImageNet-Adversarial, ImageNet-Rendition.

Due to various factors, the links to some datasets may be outdated or invalid.

To make it easy for you to download these datasets, we maintain a repository on HuggingFace, which contains all the datasets to be used (except ImageNet). Each dataset also includes the corresponding split_zhou_xx.json file.

Instructions for How to download these datasets:

Using the huggingface-cli command-line tool:

Install the CLI tool if not already installed.

pip install -U huggingface-hub

Download the datasets.

huggingface-cli download zhengli97/prompt_learning_dataset

Some projects from our lab may familiarize you with prompt learning:

Open Source Paper List: https://github.com/zhengli97/Awesome-Prompt-Adapter-Learning-for-VLMs
中文视频解读：《视觉语言模型CLIP的提示学习方法研究》，链接
Published Papers:
- Advancing Textual Prompt Learning with Anchored Attributes. ICCV 2025. [Paper] [Project Page] [Code] [中文解读] [中文翻译]
- PromptKD: Unsupervised Prompt Distillation for Vision-Language Models. CVPR 2024. [Paper] [Project Page] [Code] [中文解读] [中文翻译]
- Cascade Prompt Learning for Vision-Language Model Ddaptation. ECCV 2024. [Paper] [Code] [中文解读]
- Fine-Grained Visual Prompting. NeurIPS 2023. [Paper] [Code]