Pruna, a startup focused on compressing large AI models, has raised $3.5 million in seed funding. Their pitch deck reveals plans to shrink the size of large language models like GPT-3 by up to 90%, making them more efficient and accessible. Pruna’s compression technology aims to reduce the computational resources required to run these models, lowering costs and enabling their deployment on edge devices. The company claims its approach preserves model performance while achieving significant size reductions. Pruna’s founders, with backgrounds in machine learning and compression, believe their technology could democratize access to advanced AI capabilities. The funding round was led by Radical Ventures, with participation from investors like Liquid2 Ventures and angel investors from companies like OpenAI and Google Brain. Pruna plans to use the funding to expand its team and further develop its compression platform.
Source: https://www.businessinsider.com/pruna-ai-models-compress-funding-pitch-deck-2024-11