How to process each run? eg. example below

[Mango_man]

Hi n8ners! Loving the langchain nodes. When running a Map Reduce Summarization Chain only the final response is available as Output. How can we get each of the output for the intermediary step runs. Im seeing 30 runs in the logs and i want to add each to my supabase but only getting the final response. Thanks!

Eg in my workflow I want the 37 openai chat outputs and the 8 recursive text chunks to each be inserted to Supabase. But currently only the 1 output of the summarisation chain is.

Thanks for your help!

[MutedJam]

Hi @Mango_man, welcome to the community!

I am afraid I don’t think it’s currently possible to read data from sub-nodes like the “OpenAI Chat Model” one on your screenshot. But perhaps @oleg who worked a lot on the langchain functionality has an idea here?

[Mango_man]

Thank you for your reply.

I do hope there is a way to do this.

How can get openai to iterate through chunks and then the llm response and chunk be inserted to supabase/pinecone along with metadata?

Being able to pull the items from each run of a sub node would achieve this if possible. Or perhaps there is another workflow.

Thanks!