Model Benchmark

[Andrei-Cristian Rad]

Hi, and thanks for this repository. It's very useful!

I wanted to ask if there are any benchmarks of this approach/method (with any of the embeddings), on the literature datasets like SemEval.

Thanks :D

[Maarten Grootendorst]

At the moment, there are no benchmarks of KeyBERT due to its modular nature and derivation of some existing papers. There might be in the future but for now the focus has remained on customization instead which complicates evaluation quite a bit.

[Andrei-Cristian Rad]

Thanks for the info. Keep up the great work!