[fix] align default neuron behavior between model server and handler #2116
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
lanking520
approved these changes
Jun 27, 2024
siddvenk
approved these changes
Jun 27, 2024
tosterberg
added a commit
to tosterberg/djl-serving
that referenced
this pull request
Jun 28, 2024
…eepjavalibrary#2116) (cherry picked from commit a215752)
tosterberg
added a commit
that referenced
this pull request
Jun 28, 2024
a-ys
pushed a commit
to a-ys/djl-serving-1
that referenced
this pull request
Jun 28, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
Description
Updating the behavior of LmiConfigRecommender to match the behavior of the default properties when using a serving.properties for NeuronX containers. The recommender currently defaults all models to rolling batch, but this is not the desired behavior for LMI NeuronX due to the limited number of supported models for rolling batch.
This undoes a change that was a altered the default as aa side effect of the recommenders implementation but is not consistent in defaulting between launching with environmental variables and serving.properties for LMI NeuronX.