Improve ROSA customer experience (and customer retention) by leveraging foundation models to do “gpt-chat” style search of Red Hat customer documentation assets.
Other
24
stars
11
forks
source link
A mechanism to introspect language chain operations #30
Is your feature request related to a problem? Please describe
As we build language "chains" that involve multiple components and steps (creation of embeddings, querying embeddings, prompt template selection, API queries) I would like to be able to see what is going in detail during a query to the model.
Describe the solution you'd like
Some "debug" option for the whole chains. Possibly something that can be external so that it does not depend on being inside a notebook to be able to see it.
Is your feature request related to a problem? Please describe
As we build language "chains" that involve multiple components and steps (creation of embeddings, querying embeddings, prompt template selection, API queries) I would like to be able to see what is going in detail during a query to the model.
Describe the solution you'd like
Some "debug" option for the whole chains. Possibly something that can be external so that it does not depend on being inside a notebook to be able to see it.
Additional context
https://github.com/redhat-et/foundation-models-for-documentation/pull/15#pullrequestreview-1341421019