The news: Nvidia, the leading producer of AI hardware, just released a family of open-source AI models that are powerful enough to compete with proprietary generative AI (genAI) tools from OpenAI, Meta, and Google.
Nvidia’s NVLM 1.0 multimodal large language models (LLMs) include the flagship NVLM-D-72B, which processes complex visual and textual information and competes with OpenAI’s GPT-4 and Meta’s Llama 3.1 in math and coding applications.
Seizing the moment: Nvidia making its AI models open source and promising to release the training code could accelerate collaboration and innovation in AI research.