[Feature][MISC] Automation Workflow and Runtimes

[YouNeedCryDear]

This PR introduces comprehensive automation for container release processes and significantly expands OME's model runtime support with new Llama 3.3, Llama 4, and enhanced vision model configurations.

What type of PR is this?

/kind bug
/kind cleanup
/kind feature

What this PR does / why we need it:

• New GitHub Actions Workflow (.github/workflows/automation.yaml): 566-line
  comprehensive automation pipeline that:
  • Automatically deploys InferenceServices when containers are pushed
  • Runs performance benchmarks on deployed models
  • Executes sanity checks for model validation
  • Provides detailed deployment summaries with success/failure tracking
  • Handles cleanup of resources after testing
  • Supports both automated and manual deployment triggers

🦙 Expanded Model & Runtime Support

• Llama 3.3 70B Support: Added complete runtime configurations for both SRT
  and vLLM engines with FP8 quantization
• Llama 4 Model Family:
  • llama-4-scout-17b-16e-instruct: 4-GPU configuration with 196K context
    length
  • llama-4-maverick-17b-128e-instruct-fp8: 8-GPU configuration with 524K
    context length
  • Both support multimodal capabilities and tool calling with Pythonic
    parser
• Enhanced Vision Models: New runtimes for Llama 3.2 90B Vision with
  dynamic FP8 quantization
• Additional Model Variants: Extended support for Llama 3.1 8B and other
  configurations

Which issue(s) this PR fixes:

Fixes #

Special notes for your reviewer:

Does this PR introduce a user-facing change?

[gemini-code-assist]

Summary of Changes

Hello @YouNeedCryDear, 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 significantly enhances the OME platform by introducing a robust automation pipeline for container releases and expanding its model runtime capabilities. The changes include a new GitHub Actions workflow for automated deployment, testing, and resource management. Furthermore, it integrates support for a wider array of large language models, specifically adding new configurations for Llama 3.3, Llama 4, and improved vision models, along with various Llama 3.1 variants. These updates aim to streamline development workflows and provide broader access to advanced AI models within the OME ecosystem.

Highlights

• Comprehensive Automation Workflow: A new, extensive GitHub Actions workflow (automation.yaml) has been implemented to streamline container release processes. This automation covers deployment of InferenceServices, performance benchmarking, model validation sanity checks, detailed deployment summaries, and resource cleanup, supporting both automated and manual triggers.
• Llama 3.3 70B Model Support: Expanded support for the Llama 3.3 70B model, including configurations for both SRT (SGLang) and vLLM engines with FP8 quantization.
• New Llama 4 Model Family Integration: Introduced configurations for new Llama 4 models, specifically llama-4-scout-17b-16e-instruct (4-GPU, 196K context) and llama-4-maverick-17b-128e-instruct-fp8 (8-GPU, 524K context). Both variants now support multimodal capabilities and tool calling via a Pythonic parser.
• Enhanced Vision Model Runtimes: Enhanced vision model support with new runtimes for Llama 3.2 90B Vision, featuring dynamic FP8 quantization.
• Additional Llama Model Variants: Added extended support for various Llama 3.1 models, including the 8B variant, and other configurations, broadening the range of available models.

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. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

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 introduces a significant number of new model and runtime configurations, expanding support for various Llama models. The changes are extensive and well-structured. However, I've identified several critical and high-severity issues, mostly related to copy-paste errors in the new YAML files. These include incorrect metadata names, inconsistent model size ranges, and typos that could break deployments or cause incorrect runtime behavior. Addressing these issues is crucial for ensuring the stability and correctness of the new model configurations.

[slin1237]

Why changing this?

[YouNeedCryDear]

This is just to be consistent with what are currently in cluster. All the existing models follows this all lower case pattern.

[slin1237]

Same here
Those are just path and mimicking model ID
We don't have to change this, please revert

[YouNeedCryDear]

All lowercase or mimicking the model ID, we just need to agree on a pattern then I will clean up the rest.