This collection contains the models and datasets used in EchoLLaMA: 3D-to-Speech with Multimodal AI paper.