Google-backed Cropin launches ‘Akshara’ AI model, designed to remove knowledge barriers and empower anyone in the agriculture ecosystem to build frugal and scalable AI solutions for the sector.
The first edition of 'Akshar' will cover nine crops
Wheat, maize, sorghum, barley, cotton, sugarcane, soybean and millet for five countries of the Indian subcontinent.
Recognizing the environmental impact of running large language models (LLMs), Croppin has carefully compressed 'characters' from 16-bit to 4-bit.
“Domain-specific AI models for agriculture are expected to attract significant investment, offering a practical and economically viable approach to food system transformation,” said Krishan Kumar, Founder and CEO of Cropin.
These models could potentially transform agriculture, paving the way for a new era of technology-driven farming in a region that has traditionally seen limited technological progress.
The 'Letter' AI model was fine-tuned with over 5,000 high-quality question-answer pairs specific to agriculture and over 160,000 tokens in context.
"These numbers are expected to grow as we add more crops, geographies and use cases," the company said.
Factors such as irregular or excessive rainfall, unpredictable heat waves, increasing attacks of pests and diseases affect the functioning of farmers and reduce agricultural yield, productivity and profitability.
Croppin said it aims to bridge this gap with 'Akshar' by harnessing the power of GenAI to provide insights into modern farming practices, accurate information and agricultural advice.
The company said the open-source initiative aims to support agronomists, agronomists, field staff and extension workers and gradually expand services to farmers in multiple languages.
The first edition of 'Akshar' will cover nine crops
Wheat, maize, sorghum, barley, cotton, sugarcane, soybean and millet for five countries of the Indian subcontinent.
Recognizing the environmental impact of running large language models (LLMs), Croppin has carefully compressed 'characters' from 16-bit to 4-bit.
“Domain-specific AI models for agriculture are expected to attract significant investment, offering a practical and economically viable approach to food system transformation,” said Krishan Kumar, Founder and CEO of Cropin.
These models could potentially transform agriculture, paving the way for a new era of technology-driven farming in a region that has traditionally seen limited technological progress.
The 'Letter' AI model was fine-tuned with over 5,000 high-quality question-answer pairs specific to agriculture and over 160,000 tokens in context.
"These numbers are expected to grow as we add more crops, geographies and use cases," the company said.
Factors such as irregular or excessive rainfall, unpredictable heat waves, increasing attacks of pests and diseases affect the functioning of farmers and reduce agricultural yield, productivity and profitability.
Croppin said it aims to bridge this gap with 'Akshar' by harnessing the power of GenAI to provide insights into modern farming practices, accurate information and agricultural advice.
The company said the open-source initiative aims to support agronomists, agronomists, field staff and extension workers and gradually expand services to farmers in multiple languages.
