[Feature] Reduce the cost of torchair

[jianzs]

What this PR does / why we need it?

This can cut down on data preparation time for torchair and speed up graph dispatch.

Does this PR introduce any user-facing change?

No

How was this patch tested?

• vLLM version: v0.12.0
• vLLM main: vllm-project/vllm@ad32e3e

[gemini-code-assist]

Code Review

The pull request introduces a change to vllm_ascend/torchair/torchair_model_runner.py, specifically within the _compile_torchair_graph method. A new line torch._dynamo.config.inline_inbuilt_nn_modules = False has been added before the loop that triggers torchair graph capture. This configuration change likely aims to prevent the inlining of inbuilt neural network modules during torch.dynamo's graph compilation process. There were no review comments provided for this change.

[github-actions]

👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:‌‌

• A PR should do only one thing, smaller PRs enable faster reviews.
• Every PR should include unit tests and end-to-end tests ‌to ensure it works and is not broken by other future PRs.
• Write the commit message by fulfilling the PR description to help reviewer and future developers understand.

If CI fails, you can run linting and testing checks locally according Contributing and Testing.