[vov_or]

Yes, we are training text embedding models right now. And also have plans to open-source some of them! In addition, we train encoders for different modalities with retrieval purposes. For example, video data.