RAG (DPR+seq2seq) pre-trained example

I’d appreciate if someone can post here a few lines of code that show the full usage of a pre-trained RAG model with the combination of DPR and seq generation (not token prediction) that includes pre-trained models that were trained jointly and do not require any fine-tuning.

I got some error messages with using a config file and hence I’d appreciate if someone can help me out. TIA!!