Nvidia Unveils NVLM 1.0, a Powerful Open-Source AI Model Rivalling GPT-4

The new multimodal language model NVLM-D-72B excels in both vision and text tasks, with Nvidia making its model weights and training code publicly available.

Nvidia's NVLM 1.0 family includes the flagship NVLM-D-72B model, boasting 72 billion parameters.
NVLM-D-72B demonstrates state-of-the-art performance in vision-language tasks and improves text-only accuracy.
The model can interpret memes, analyze images, solve math problems, and code software.
Nvidia's decision to open-source the model weights and training code is unprecedented among AI giants.
The release is expected to accelerate AI research and development, benefiting smaller firms and independent researchers.