[docs] updates the serving runtime documentation

[slin1237]

What type of PR is this?

/kind documentation

What this PR does / why we need it:

This PR updates the serving runtime documentation to accurately reflect the current runtime configurations available in the codebase. The documentation for both SRT
(SGLang) and VLLM runtimes was outdated, showing runtimes that no longer exist and missing ones that have been added.

The documentation now correctly lists all available runtimes in /config/runtimes/srt/ and /config/runtimes/vllm/.

Which issue(s) this PR fixes:

Fixes #

Special notes for your reviewer:

SRT Runtime Updates:

• Updated to reflect all 21 available runtimes
• Added missing PD (prefill-decode) runtime variants
• Fixed naming consistency by adding -rt suffix to all runtime names
• Removed non-existent srt-llama-3-1-405b-instruct-fp8 runtime
• Fixed typo: srt-mmixtral-8x7b-instruct → mixtral-8x7b-instruct-rt

VLLM Runtime Updates:

• Updated to reflect all 14 available runtimes
• Removed 5 non-existent runtimes:
  • vllm-ray-multi-node-deepseek-v3-rdma
  • vllm-llama-3-1-70b-instruct
  • vllm-llama-3-3-70b-instruct-fp8-dynamic
  • vllm-llama-3-70b-instruct
  • vllm-phi-3-vision-128k-instruct
• Fixed naming consistency (added -rt suffix, removed vllm- prefix)

Does this PR introduce a user-facing change?

Updated serving runtime documentation to accurately reflect available SRT and VLLM runtimes, removing outdated entries and adding missing runtime configurations

[gemini-code-assist]

Summary of Changes

Hello @slin1237, 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!

This pull request focuses on updating the documentation for serving runtimes, specifically for SRT (SGLang) and VLLM, to ensure it accurately reflects the currently available configurations. The changes involve synchronizing the documented runtime names with the codebase, removing outdated entries, and adding newly supported variants.

Highlights

• SRT Runtime Documentation Update: The documentation for SGLang (SRT) runtimes has been comprehensively updated to list all 21 available configurations. This includes adding missing prefill-decode (PD) variants, ensuring consistent naming with an -rt suffix, removing an outdated runtime entry, and correcting a typo.
• VLLM Runtime Documentation Update: The VLLM runtime documentation has been revised to reflect the 14 currently available runtimes. This involved removing several non-existent entries and standardizing naming conventions by adding an -rt suffix and removing the vllm- prefix.
• General Documentation Consistency: A minor update was made to a navigation link in the concepts index, changing 'Benchmarks' to 'Benchmark' for improved consistency.

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 is currently in preview and 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 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 serving runtime documentation for SRT and VLLM to match the available runtimes in the codebase. The changes are well-described and accurately reflect the PR's intent. There are a couple of minor sorting issues in the runtime tables that could improve readability.

[gemini-code-assist]

The list of llama-3-2 runtimes is not sorted in a natural order, which can be confusing for readers. For example, 11b is listed before 1b. Reorder these runtimes based on the model size for better readability.

Suggested change

	| llama-3-2-11b-vision-instruct-rt | transformers | safetensors | MllamaForConditionalGeneration |
	| llama-3-2-1b-instruct-pd-rt | transformers | safetensors | LlamaForCausalLM |
	| llama-3-2-1b-instruct-rt | transformers | safetensors | LlamaForCausalLM |
	| llama-3-2-3b-instruct-pd-rt | transformers | safetensors | LlamaForCausalLM |
	| llama-3-2-3b-instruct-rt | transformers | safetensors | LlamaForCausalLM |
	| llama-3-2-90b-vision-instruct-rt | transformers | safetensors | MllamaForConditionalGeneration |
	| llama-3-2-1b-instruct-pd-rt | transformers | safetensors | LlamaForCausalLM |
	| llama-3-2-1b-instruct-rt | transformers | safetensors | LlamaForCausalLM |
	| llama-3-2-3b-instruct-pd-rt | transformers | safetensors | LlamaForCausalLM |
	| llama-3-2-3b-instruct-rt | transformers | safetensors | LlamaForCausalLM |
	| llama-3-2-11b-vision-instruct-rt | transformers | safetensors | MllamaForConditionalGeneration |
	| llama-3-2-90b-vision-instruct-rt | transformers | safetensors | MllamaForConditionalGeneration |