ImageBind fine-tuning with LoRA

Follow the full discussion on Reddit.
ImageBind is a novel multimodal neural network that can learn a universal representation for various types of data, such as images, videos, audio, text, IMU data, and heat maps. It uses large-scale pre-trained models and contrastive learning to achieve this. If you want to fine-tune ImageBind for your own task, you can use ImageBind-LoRA, which applies Low-Rank Adaptation (LoRA) to adjust the embeddings.