KurmaAI/AQUA-7B
KurmaAI/AQUA-7B is a 7-billion parameter large language model, fine-tuned from Mistral 7B v0.3, developed by Kurma AI. It is specifically designed for the global aquaculture industry, providing actionable insights across species-specific farming, hatchery operations, water quality control, and disease management. Trained on over 3 million real and synthetic aquaculture conversations (approximately 1 billion tokens), AQUA-7B excels at domain-specific tasks for fish farms, hatcheries, and aquaculture researchers.
Loading preview...
AQUA-7B: The First LLM for Aquaculture
AQUA-7B, developed by Kurma AI, is a 7-billion parameter large language model built exclusively for the global aquaculture industry. Fine-tuned from Mistral 7B v0.3, it is the first LLM dedicated to this domain, providing specialized insights for various aquaculture operations.
Key Capabilities
- Comprehensive Aquaculture Management: Covers production systems (ponds, tanks, cages, RAS), species management (tilapia, salmon, shrimp, etc.), genetics, hatchery operations, and early life stage care.
- Operational Guidance: Offers protocols for nutrition, feeding, growth optimization, water quality control (temperature, oxygen, pH, ammonia), and structured disease management (identification, vaccination, biosecurity).
- Sustainability & Innovation: Promotes eco-friendly practices, waste management, biodiversity, climate adaptation, and guides the adoption of new technologies like AI, automation, and sensors.
- Business & Market Insights: Advises on market trends, business planning, regulation, certification, traceability, and best practices for harvesting, processing, and food safety.
Training and Performance
AQUA-7B was trained on approximately 3 million real and synthetic Q&A pairs, totaling around 1 billion tokens of high-quality, domain-specific data. This dataset includes extension worker–farmer dialogues, FAO/ICAR/NOAA research, and synthetic Q&A from over 5,000 aquaculture topics, carefully curated for species-specific culture methods. The model was fine-tuned using a LoRA-based Supervised Fine-Tuning (SFT) approach on 16 NVIDIA H200 GPU Multi Clusters.
Good for
- Generating insights for fish farms and hatcheries.
- Assisting aquaculture researchers and Aqua-Tech innovators.
- Providing guidance on species-specific farming practices.
- Supporting decision-making in water quality, disease management, and sustainable aquaculture.