Model comparison for question pairs detection using 10-fold cross validation.

<p>Model comparison for question pairs detection using 10-fold cross validation.</p>

Saved in:

Bibliographic Details
Main Author:	Sifei Han (3747112) (author)
Other Authors:	Lingyun Shi (3907438) (author), Fuchiang (Rich) Tsui (20542907) (author)
Published:	2025
Subjects:	Science Policy Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Information Systems not elsewhere classified text similarity tasks supervised model training quora question pairs natural language processing identifying potential reviewers 72 %- 65 13 %- 85 supervised learning using tuned llama model 83 %- 82 4 %) outperformed 25 %]; p 9 %) using 7b model using recent larger llm larger pretrained models improved performance compared conducted various fine improved performance using 100 larger llm shot learning tuning approaches study demonstrated smaller llm secondary study recommending subject rank adaptation qqp ), prompt engineering previous study numerous datasets nlp ). matter experts matching resumes lora ]), like chatgpt job descriptions improvement compared grant proposals glue benchmark fold cross fair comparison case study best results 7b parameters 67 %), 20 %).
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items