SGLang Integration + Accuracy Tests, Restructure app_tests/integration_tests #570
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
renxida
approved these changes
Nov 19, 2024
Restructure app_tests/integration_tests, Add copyright headers to files in integration_tests that were missing it
Add more logging and a little cleanup in sglang_frontend_test
renxida
force-pushed
the
slg-integration-tests
branch
from
6f56b59 to
627af2d
Compare
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Outdated suggestions cannot be applied.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.
Description
This PR implements integration tests for the Shortfin LLM Server w/ the SGLang integration. It uses llama3-8b-instruct on GPU, which is downloaded using sharktank's
hf_datasetsscript.The tests server two purposes:
1.0, so the responses should be deterministic.This test is intended to run every 4 hours, which allows for us to detect degradations in shortfin LLM output accuracy. If we do get a failure due to an accuracy degradation, there will only be a small set of shark-ai/iree commits that could be responsible.