[QA] MoMa: Efficient Early-Fusion Pre-training with Mixture of Modality-Aware Experts

Arxiv Papers - [QA] MoMa: Efficient Early-Fusion Pre-training with Mixture of Modality-Aware Experts