Is there still no way to perform batched parallel Q&A predictions, using a HuggingFace pipeline? It seems to be possible with the other pipelines but not Q&A.
Any advice on how to do this?
1 Like
Is there still no way to perform batched parallel Q&A predictions, using a HuggingFace pipeline? It seems to be possible with the other pipelines but not Q&A.
Any advice on how to do this?