Allow overrides of benchmark test configs - #146
Open
sfc-gh-mwyatt wants to merge 4 commits into
Open
Conversation
sfc-gh-mwyatt
requested review from
sfc-gh-aqiao,
sfc-gh-jrasley,
sfc-gh-mhidayetoglu and
sfc-gh-yewang
as code owners
July 23, 2025 22:06
sfc-gh-aqiao
approved these changes
Jul 23, 2025
sfc-gh-yewang
approved these changes
Jul 24, 2025
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Added
--benchmark-configoption when running pytest to override defaults:❯ python -m pytest --collect-only -q tests/benchmarks/test_benchmarks.py::test_performance[llama_8b-batch] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b-single] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b-arc_challenge] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b-gsm8k] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b-ifeval] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b-mmlu_pro] tests/benchmarks/test_benchmarks.py::test_json_mode[llama_8b-json_mode_score] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_shift-batch] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_shift-single] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_shift-arc_challenge] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_shift-gsm8k] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_shift-ifeval] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_shift-mmlu_pro] tests/benchmarks/test_benchmarks.py::test_json_mode[llama_8b_shift-json_mode_score] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_swiftkv-batch] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_swiftkv-single] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_swiftkv-arc_challenge] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_swiftkv-gsm8k] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_swiftkv-ifeval] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_swiftkv-mmlu_pro] tests/benchmarks/test_benchmarks.py::test_json_mode[llama_8b_swiftkv-json_mode_score] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_suffix-batch] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_suffix-single] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_suffix-arc_challenge] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_suffix-gsm8k] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_suffix-ifeval] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_suffix-mmlu_pro] tests/benchmarks/test_benchmarks.py::test_json_mode[llama_8b_suffix-json_mode_score] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_spec-batch] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_spec-single] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_spec-arc_challenge] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_spec-gsm8k] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_spec-ifeval] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_spec-mmlu_pro] tests/benchmarks/test_benchmarks.py::test_json_mode[llama_8b_spec-json_mode_score] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_all-batch] tests/benchmarks/test_benchmarks.py::test_performance[llama_8b_all-single] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_all-arc_challenge] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_all-gsm8k] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_all-ifeval] tests/benchmarks/test_benchmarks.py::test_accuracy[llama_8b_all-mmlu_pro] tests/benchmarks/test_benchmarks.py::test_json_mode[llama_8b_all-json_mode_score] 42 tests collected in 0.06s❯ python -m pytest --collect-only -q --benchmark-config ./example_config_override.yaml tests/benchmarks/test_benchmarks.py::test_performance[my_custom_config-my_batch_test] tests/benchmarks/test_benchmarks.py::test_accuracy[my_custom_config-my_arc_test] tests/benchmarks/test_benchmarks.py::test_json_mode[my_custom_config-my_json_test] 3 tests collected in 0.07sTODO: