Gentoo Packages

Search

Install

Install this package:

emerge -a dev-python/vllm

If the package is masked, you can unmask it using the autounmask tool or standard emerge options:

autounmask dev-python/vllm

Or alternatively:

emerge --autounmask-write -a dev-python/vllm

Package Information

Description:

vLLM is a fast and easy-to-use library for LLM inference and serving. Provides a Python API and an OpenAI-compatible HTTP server. USE=cpu / cuda / rocm pick a single VLLM_TARGET_DEVICE for the build (mutually exclusive). Default (none of the three) builds with VLLM_TARGET_DEVICE=empty — Python entrypoints import cleanly, backend kernels fail at first model load. Useful if you only want the API surface for development.

Homepage:

https://github.com/vllm-project/vllm https://docs.vllm.ai/ https://pypi.org/project/vllm/

License:

Apache-2.0

Versions

Metadata

Description

• vLLM is a fast and easy-to-use library for LLM inference and serving. Provides a Python API and an OpenAI-compatible HTTP server. USE=cpu / cuda / rocm pick a single VLLM_TARGET_DEVICE for the build (mutually exclusive). Default (none of the three) builds with VLLM_TARGET_DEVICE=empty — Python entrypoints import cleanly, backend kernels fail at first model load. Useful if you only want the API surface for development.

Maintainers

• Ivan S. Titov (iohann.s.titov@gmail.com) - Type: person

Upstream

• Bugs: https://github.com/vllm-project/vllm/issues
• Remote IDs:
  • pypi: vllm
  • github: vllm-project/vllm

<pkgmetadata>
	<maintainer type="person">
		<email>iohann.s.titov@gmail.com</email>
		<name>Ivan S. Titov</name>
	</maintainer>
	<longdescription lang="en">
    vLLM is a fast and easy-to-use library for LLM inference and
    serving. Provides a Python API and an OpenAI-compatible HTTP server.

    USE=cpu / cuda / rocm pick a single VLLM_TARGET_DEVICE for the
    build (mutually exclusive). Default (none of the three) builds
    with VLLM_TARGET_DEVICE=empty — Python entrypoints import cleanly,
    backend kernels fail at first model load. Useful if you only want
    the API surface for development.
  </longdescription>
	<use>
		<flag name="cpu">Build for CPU inference (VLLM_TARGET_DEVICE=cpu); pull torchaudio + numba</flag>
		<flag name="rocm">Build for AMD ROCm inference (VLLM_TARGET_DEVICE=rocm); pull HIP libs + torch{audio,vision}</flag>
	</use>
	<upstream>
		<bugs-to>https://github.com/vllm-project/vllm/issues</bugs-to>
		<remote-id type="pypi">vllm</remote-id>
		<remote-id type="github">vllm-project/vllm</remote-id>
	</upstream>
</pkgmetadata>

Lint Warnings

• [Warning] USE flag 'cuda' in ebuild 0.20.1 is not documented in metadata.xml
• [Error] Missing md5-cache for ebuild vllm-0.20.1

USE Flags

Manage flags for this package: euse -i <flag> -p dev-python/vllm | euse -E <flag> -p dev-python/vllm | euse -D <flag> -p dev-python/vllm

Files

• vllm-0.20.1-cpu-system-libgomp.patch

Manifest