Leveraging Synthetic Data for Question Answering with Multilingual LLMs in the Agricultural Domain
By: Rishemjit Kaur , Arshdeep Singh Bhankhar , Jashanpreet Singh Salh and more
Potential Business Impact:
Helps farmers get farming advice in their language.
Enabling farmers to access accurate agriculture-related information in their native languages in a timely manner is crucial for the success of the agriculture field. Publicly available general-purpose Large Language Models (LLMs) typically offer generic agriculture advisories, lacking precision in local and multilingual contexts. Our study addresses this limitation by generating multilingual (English, Hindi, Punjabi) synthetic datasets from agriculture-specific documents from India and fine-tuning LLMs for the task of question answering (QA). Evaluation on human-created datasets demonstrates significant improvements in factuality, relevance, and agricultural consensus for the fine-tuned LLMs compared to the baseline counterparts.
Similar Papers
Leveraging Synthetic Data for Question Answering with Multilingual LLMs in the Agricultural Domain
Computation and Language
Helps farmers get farming advice in their language.
Evaluating Small Language Models for Agentic On-Farm Decision Support Systems
Artificial Intelligence
Lets AI help farmers on their own computers.
Unlocking the Potential of Large Language Models in the Nuclear Industry with Synthetic Data
Computation and Language
Makes nuclear information usable for smart computer programs.