feat: Implement PipelinePair primitive and related utility functions
#4
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
davidsbatista
approved these changes
May 17, 2024
|
This was a rather a big one, I tried to cover as much detail as possible, the logic looks OK to me, but I might have missed some edge cases - maybe checking what the test coverage shows to be sure we are testing every relevant piece of code. |
julian-risch
approved these changes
May 21, 2024
Pull Request Test Coverage Report for Build 9189096904Details
💛 - Coveralls |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is the first of a series of PRs that revolves around implementing an evaluation harness (c.f deepset-ai/haystack#7526).
PipelinePairis a utility class that allows for two pipelines to be used in tandem. The motivation behind this abstraction is to reduce the amount of glue code it takes to connect a, say, RAG pipeline with an evaluation pipeline. It aims to be generic enough work with arbitrary pipelines while making special accommodations to ensure that pipelines of different "shapes" can be connected without much extra work from the user. "Shape" here refers to the modality of pipeline inputs and outputs, e.g: a typical RAG pipeline takes a query and returns an answer whereas a typical RAG evaluation pipeline will take batches of queries and answers. Connecting two such pipelines is handled by automatically aggregating and deaggregating inputs and outputs at the interface between the pipelines.Blocked by #2.