Introduce support for dynamic batching #1662
Draft
Conversation
This file contains hidden or 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
…into baijumeswani/engine
…into baijumeswani/engine
…into baijumeswani/engine
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Outdated suggestions cannot be applied.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.
The changes in #1580 introduce a new API in
onnxruntime-genaithat enables continuous submission of requests to a generation engine. Internally, the scheduler statically batches these requests, allowing the API to interface seamlessly with models designed for static batching.This pull request lays the groundwork for dynamic request batching by incorporating concepts from vLLM. Specifically, it implements a block-based key-value cache manager.
To support dynamic batching, the underlying model must handle variable-length inputs (as opposed to fixed-size static batches) and leverage the
PagedAttentioncontrib-op.