chore: use torch 2.8 stable

[zhyncs]

Motivation

Modifications

Accuracy Test

Benchmark & Profiling

Checklist

• Format your code according to the Code Formatting with Pre-Commit.
• Add unit tests as outlined in the Running Unit Tests.
• Update documentation / docstrings / example tutorials as needed, according to Writing Documentation.
• Provide throughput / latency benchmark results and accuracy evaluation results as needed, according to Benchmark and Profiling and Accuracy Results.
• For reviewers: If you haven't made any contributions to this PR and are only assisting with merging the main branch, please remove yourself as a co-author when merging the PR.
• Please feel free to join our Slack channel at https://slack.sglang.ai to discuss your PR.

[gemini-code-assist]

Summary of Changes

Hello @zhyncs, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

I've made a chore change to update our project's dependency on PyTorch. The core purpose of this pull request is to transition from using PyTorch's test or nightly builds to the stable 2.8 release. This ensures that our development and CI environments are running on a more robust and officially supported version of the framework, which should lead to more predictable and reliable behavior.

Highlights

• Dependency Update: I've updated the PyTorch installation URLs across the Dockerfile, CI scripts, and kernel build scripts to point to the stable PyTorch 2.8 release instead of the previous 'test' or nightly builds.
• Build Stability: This change aims to enhance the overall stability and reliability of our builds and development environment by leveraging a stable and officially released version of PyTorch.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments or fill out our survey to provide feedback.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request updates the build and CI scripts to use the stable PyTorch 2.8 wheels by removing /test from the download URLs. The changes are correct and align with the goal of using a stable release. I've suggested a small refactoring in sgl-kernel/build.sh to improve code maintainability by reducing duplication.

[gemini-code-assist]

The if/elif/else block contains duplicated code for setting DOCKER_IMAGE and constructing the TORCH_INSTALL command. You can refactor this to be more concise and maintainable by setting DOCKER_IMAGE once and using a case statement to determine the CUDA version suffix for the PyTorch URL. This will make the script easier to read and modify in the future.

Suggested change

	if [ ${CUDA_VERSION} = "12.9" ]; then
	DOCKER_IMAGE="${BUILDER_NAME}:cuda${CUDA_VERSION}"
	TORCH_INSTALL="pip install --no-cache-dir torch==2.8.0 --index-url https://download.pytorch.org/whl/test/cu129"
	TORCH_INSTALL="pip install --no-cache-dir torch==2.8.0 --index-url https://download.pytorch.org/whl/cu129"
	elif [ ${CUDA_VERSION} = "12.8" ]; then
	DOCKER_IMAGE="${BUILDER_NAME}:cuda${CUDA_VERSION}"
	TORCH_INSTALL="pip install --no-cache-dir torch==2.8.0 --index-url https://download.pytorch.org/whl/test/cu128"
	TORCH_INSTALL="pip install --no-cache-dir torch==2.8.0 --index-url https://download.pytorch.org/whl/cu128"
	else
	DOCKER_IMAGE="${BUILDER_NAME}:cuda${CUDA_VERSION}"
	TORCH_INSTALL="pip install --no-cache-dir torch==2.8.0 --index-url https://download.pytorch.org/whl/test/cu126"
	TORCH_INSTALL="pip install --no-cache-dir torch==2.8.0 --index-url https://download.pytorch.org/whl/cu126"
	fi
	DOCKER_IMAGE="${BUILDER_NAME}:cuda${CUDA_VERSION}"
	case "${CUDA_VERSION}" in
	"12.9") CU_SUFFIX="129" ;;
	"12.8") CU_SUFFIX="128" ;;
	*) CU_SUFFIX="126" ;;
	esac
	TORCH_INSTALL="pip install --no-cache-dir torch==2.8.0 --index-url https://download.pytorch.org/whl/cu${CU_SUFFIX}"