[Question] How to extend implicit for image and text or multimodal recommendations using Transformer models as encoders ?

[PrithivirajDamodaran]

Please advice.

[benfred]

This isn't something that is explicitly supported here - you should look for other deep learning recommender toolkits for this. I'd suggest something like https://github.com/nvidia-merlin/models or https://github.com/pytorch/torchrec maybe.