Benchmark For Few Shot Classification with Many Classes
FastFit
non-profit
AI & ML interests
None defined yet.
Organization Card
About org cards
FastFit, a method, and a Python package design to provide fast and accurate few-shot classification, especially for scenarios with many semantically similar classes. FastFit utilizes a novel approach integrating batch contrastive learning and token-level similarity score. Compared to existing few-shot learning packages, such as SetFit, Transformers, or few-shot prompting of large language models via API calls, FastFit significantly improves multi-class classification performance in speed and accuracy across FewMany, our newly curated English benchmark, and Multilingual datasets. FastFit demonstrates a 3-20x improvement in training speed, completing training in just a few seconds.
pip install fast-fit
Collections
1
models
None public yet
datasets
14
FastFit/hwu_64
Viewer
•
Updated
•
896
FastFit/clinc_150
Viewer
•
Updated
•
906
FastFit/banking_77
Viewer
•
Updated
•
1.13k
•
3
FastFit/massive_es_60
Viewer
•
Updated
FastFit/massive_fr_60
Viewer
•
Updated
FastFit/massive_zh_60
Viewer
•
Updated
FastFit/massive_ja_60
Viewer
•
Updated
FastFit/massive_de_60
Viewer
•
Updated
FastFit/massive_en_60
Viewer
•
Updated
FastFit/argument_topic_71
Viewer
•
Updated
•
899
•
1