Clean up remaining Punica C information

[jeejeelee]

#5036 has been merged. Cleaning up the remaining Punica C information.

@mgoin I see that most of these formatting-related files were contributed by you. Could you please take a look at this PR? Thanks~

[github-actions]

👋 Hi! Thank you for contributing to the vLLM project.
Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which consists a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of default ones by unblocking the steps in your fast-check build on Buildkite UI.

Once the PR is approved and ready to go, please make sure to run full CI as it is required to merge (or just use auto-merge).

To run full CI, you can do one of these:

• Comment /ready on the PR
• Add ready label to the PR
• Enable auto-merge.

🚀

[mgoin]

Thanks, this looks reasonable to me. Let me take a look myself in the codebase

[simon-mo]

I recently comment out the LoRA test to save CI cost. Please run those locally and report back. 🙇‍♂️

[simon-mo]

You can also uncomment them in .buildkite/test-pipeline.yaml and run it in CI

[jeejeelee]

@simon-mo I will considering optimizing the LoRA-related unit tests . Additionally, regarding this PR, what test content should I provide feedback on?

[simon-mo]

vllm/.buildkite/test-pipeline.yaml

Lines 204 to 217 in c8a7e93

	# - label: LoRA Test %N
	# #mirror_hardwares: [amd]
	# command: pytest -v -s lora --shard-id=$$BUILDKITE_PARALLEL_JOB --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --ignore=lora/test_long_context.py
	# parallelism: 4
	# - label: LoRA Long Context (Distributed)
	# #mirror_hardwares: [amd]
	# num_gpus: 4
	# # This test runs llama 13B, so it is required to run on 4 GPUs.
	# commands:
	# # FIXIT: find out which code initialize cuda before running the test
	# # before the fix, we need to use spawn to test it
	# - export VLLM_WORKER_MULTIPROC_METHOD=spawn
	# - pytest -v -s -x lora/test_long_context.py

For this PR to merge, please uncomment these and let it pass in CI first. We are working on optimizing the steps overall. Optimizing the LoRA test is definitely welcomed!

[jeejeelee]

@simon-mo The LoRA tests have all passed. It seems that a force merge may be necessary. Furthermore, I've noticed that the test for each LoRA slice takes approximately 20 minutes. If we were to optimize the LoRA testing process, should we focus on reducing this time consumption or on improving the resource efficiency within the LoRA tests?

[simon-mo]

Great. Can you comment it out for now so it doesn't trigger every commit. And we can force merge

[jeejeelee]

Great. Can you comment it out for now so it doesn't trigger every commit. And we can force merge

Done.

[WoosukKwon]

LGTM! Let's merge this PR? @jeejeelee @simon-mo

[jeejeelee]

LGTM! Let's merge this PR? @jeejeelee @simon-mo

Agreed, I think this PR's task is complete, and all related tests have passed