airsplay/lxmert

Features

• Multimodal Pretraining - Learning cross-modality encoder representations from transformers.
• Vision Language Models - Cross-modality encoder representations learned via transformers.

Star history