There is currently no option in ingestion pipelines to use chunked inferences. This story is to implement the ability to do chunked inference within ingestion pipelines. The team has not yet aligned on whether we want to enable chunking in the existing ingestion pipeline processor or in a new processor? This will require some investigation to understand what the complexity/UX implications of either option are.
Description
Description
There is currently no option in ingestion pipelines to use chunked inferences. This story is to implement the ability to do chunked inference within ingestion pipelines. The team has not yet aligned on whether we want to enable chunking in the existing ingestion pipeline processor or in a new processor? This will require some investigation to understand what the complexity/UX implications of either option are.