Bengaluru, April 16 Agritech company Cropin Technology on Tuesday launched the sector's first open-sourced micro-language model for climate-smart agriculture, targeting undeserved farmers in Global South.
The Google-backed Cropin launched 'akṣara' AI model which is designed to remove barriers to knowledge, and empower anyone in the agriculture ecosystem to build frugal and scalable AI solutions for the sector.
The first version of 'akṣara' will cover nine crops -- paddy, wheat, maize, sorghum, barley, cotton, sugarcane, soybean, and millets for five countries in the Indian subcontinent.
Recognising the environmental impact of running large language models (LLMs), Cropin has meticulously compressed 'akṣara' into 4-bit from 16-bit.
"Domain-specific AI models for agriculture are expected to attract significant investments, offering a practical and economically viable approach to food systems transformation," said Krishna Kumar, Founder and CEO, Cropin.
These models can potentially transform agriculture, paving the way for a new era of tech-driven farming in a sector that has traditionally seen limited technological advancement.
The 'akṣara' AI model was fine-tuned with more than 5,000 high-quality question-response pairs specific to agriculture and more than 160,000 tokens in the context.
"These numbers are expected to increase as we add more crops, geographic locations and use cases," said the company.
Factors like irregular or extreme rainfall, unpredictable heatwaves, and increased pest and disease attacks affect farmers' practices and reduce agricultural yield, productivity, and profitability.
Cropin said it aims to bridge this gap with ‘akṣara’ by harnessing the power of GenAI to provide insights into modern farming practices, accurate information, and farm advisories.
The open-source initiative aims to support agronomists, agri-scientists, field staff, and extension workers and gradually extend the services to farmers in multiple languages, the company added.
Disclaimer: This post has been auto-published from an agency feed without any modifications to the text and has not been reviewed by an editor