Newsletter
Join the Community
Subscribe to our newsletter for the latest news and updates
Train a 26M-parameter visual multimodal VLM from scratch in just 1 hour, suitable for deep learning enthusiasts.
MiniMind-V is an innovative visual language model (VLM) that allows you to train a 26M-parameter model from scratch in just 1 hour using a single NVIDIA 3090 GPU. This project aims to provide a minimal and effective implementation of VLMs, emphasizing accessibility for individuals with basic hardware setups.
Join the MiniMind-V project to explore the fascinating world of visual language models and contribute to its development!