[docker] Restructure Dockerfile for more efficient and cache-friendly builds

[amrmahdi]

Approach:

• Pre-install PyTorch, FlashInfer, and other slow-changing dependencies in vllm-base before installing the vLLM wheel for better layer caching
• Add parallel extensions-build stage for DeepGEMM and EP kernels
• Move stable packages (accelerate, bitsandbytes, etc.) earlier in build

This allows incremental builds with Python-only changes to skip the expensive dependency installation layers.

Performance:

Incremental builds with Python-only changes now complete in ~16 minutes (previously 35+ minutes).

Future work:

We considered building these base stages as separate images that could be built independently and baked into CI AMIs for maximum cache reuse. However, this introduces maintenance burden of extra pipelines and update strategies. The inline approach is simpler and can be optimized later by baking the main build image into AMIs daily to maximize layer cache reuse.

[chatgpt-codex-connector]

Codex usage limits have been reached for code reviews. Please check with the admins of this repo to increase the limits by adding credits.

[gemini-code-assist]

Code Review

This pull request significantly refactors the Dockerfile to improve build efficiency and leverage caching more effectively. The introduction of a parallel extensions-build stage and the restructuring of the vllm-base stage to pre-install slow-changing dependencies are excellent changes that should substantially reduce incremental build times. The overall approach is well-thought-out and correctly implemented. I have one suggestion to further optimize Docker image layering, but the pull request is a great improvement overall.

[mergify]

Hi @amrmahdi, the pre-commit checks have failed. Please run:

uv pip install pre-commit
pre-commit install
pre-commit run --all-files

Then, commit the changes and push to your branch.

For future commits, pre-commit will run automatically on changed files before each commit.

Tip

Is mypy or markdownlint failing?

mypy and markdownlint are run differently in CI. If the failure is related to either of these checks, please use the following commands to run them locally:

# For mypy (substitute "3.10" with the failing version if needed)
pre-commit run --hook-stage manual mypy-3.10
# For markdownlint
pre-commit run --hook-stage manual markdownlint

[mergify]

Documentation preview: https://vllm--30626.org.readthedocs.build/en/30626/

[amrmahdi]

As a datapoint, my rebase build took 16m20s https://buildkite.com/vllm/ci/builds/43561/steps/canvas?sid=019b207f-7cc2-4327-8035-07fa0e925428

[mgoin]

LGTM overall, great work. Just a few quick nits to cleanup

[mgoin]

LGTM, thanks for the nice work!