FSL-Mate is a collection of resources for few-shot learning (FSL).
In particular, FSL-Mate currently contains
- FewShotPapers: a paper list which tracks the research advances on FSL
- PaddleFSL: a PaddlePaddle-based python library for FSL
We are endeavored to constantly update FSL-Mate. Hopefully, it can make FSL easier.
-
[2022-04-02] PaddleFSL v1.0.1 is officially released! It now supports bioinformatics applications and diverse few-shot natural language understanding (FewNLU) tasks. Take a look at here!
-
[2022-03-31] Add FSL papers published in ICLR 2022.
-
[2022-03-08] FSL-Mate collaborates with PaddlePaddle Hackathon and offer three missions to community users: 91. Add random search, 92. Add TPE hyperparameter search, and 93. Reorganize MAML and ANIL. Solutions which complete the missions and pass the evaluation will obtain case prizes: 1000 yuan for mission 91, 5000 yuan for mission 92, and 10000 yuan for mission 93. Come and join the Hackathon here!
Please cite our paper if you find it helpful.
@article{wang2020generalizing,
title={Generalizing from a few examples: A survey on few-shot learning},
author={Wang, Yaqing and Yao, Quanming and Kwok, James T and Ni, Lionel M},
journal={ACM Computing Surveys},
volume={53},
number={3},
pages={1--34},
year={2020},
publisher={ACM New York, NY, USA}
}
We welcome advices and feedbacks for FSL-Mate. Please feel free to open an issue or contact Yaqing Wang.