[Doc] Adding models/pipelines/features Tutorial

[wtomin]

Comments are welcomed! Suggested changes are welcomed!

Purpose

It is important to have clear, easy-to-follow, tutorials on how to adapt huggingface models/pipelines to vLLM-Omni, and support various features:

• torch.compile
• CFG Parallel
• Tensor Parallel
• Sequence Parallel
• Cache acceleration: cache-dit and TeaCache
• Quantization (coming soon, not included in this PR)
• Patch VAE Parallel (coming soon, not included in this PR)

The HowToAdd tutorial should at least cover the following content:

# How to add [Feature Name] support for a new model

## Table of Contents

## Overview

### What is xx?
Concept explanation

### Architecture
Key APIs required for adapation

## Step-by-Step Implementation
## Customization (if required customized adaptation for special models)
Or for features with multiple implementation approaches:
## Approach 1: [Method Name]
## Approach 2: [Alternative Method Name]


## Testing
How to test the performance and quality

## Troubleshooting
Common issues and solutions

## Reference Implementations
Complete examples 

## Summary

---

Essential Elements of an Effective PR Description Checklist

• The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
• The test plan, such as providing test command.
• The test results, such as pasting the results comparison before and after, or e2e results
• (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.
• (Optional) Release notes update. If your change is user facing, please update the release notes draft.

BEFORE SUBMITTING, PLEASE READ https://github.com/vllm-project/vllm-omni/blob/main/CONTRIBUTING.md (anything written below this line will be removed by GitHub Actions)

[dongbo910220]

This is a much-needed tutorial. Excited to see this land.

[Copilot]

Pull request overview

This PR adds comprehensive tutorial documentation for adapting HuggingFace models/pipelines to vLLM-Omni and supporting various advanced features including parallelism strategies and cache acceleration.

Changes:

• Reorganized parallelism documentation by moving detailed SP/CFG-Parallel content from parallelism_acceleration.md to dedicated feature-specific guides
• Completely rewrote adding_diffusion_model.md with step-by-step instructions, examples, and troubleshooting
• Added five new feature tutorial documents covering tensor parallel, CFG parallel, sequence parallel, TeaCache, and Cache-DiT
• Updated navigation structure to include new "Advanced Features" section

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 14 comments.

Show a summary per file

File	Description
docs/user_guide/diffusion/parallelism_acceleration.md	Removed detailed implementation guides (moved to dedicated feature docs)
docs/contributing/model/adding_diffusion_model.md	Complete rewrite with comprehensive step-by-step guide for adding diffusion models
docs/contributing/features/tensor_parallel.md	New guide for adding Tensor Parallel support to transformers
docs/contributing/features/teacache.md	New guide for adding TeaCache acceleration support
docs/contributing/features/sequence_parallel.md	New guide for adding Sequence Parallel support (moved from parallelism_acceleration.md)
docs/contributing/features/cfg_parallel.md	New guide for adding CFG-Parallel support to pipelines (moved from parallelism_acceleration.md)
docs/contributing/features/cache_dit.md	New guide for adding Cache-DiT acceleration support
docs/.nav.yml	Added "Advanced Features" navigation section with links to new feature guides

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[fhfuih]

Add example and documentation? executable Python examples are at examples/(offline or online)/(task and modality)/*.{py|sh}, and documentations are at examples/(offline or online)/(task and modality)/*.md and docs/user_guide/examples/(offline or online)/*.md

[fhfuih]

This is different from the supported_models and ...acceleration markdowns you have added below.

[wtomin]

@dongbo910220 Comments are welcomed! I think you maybe interested in sp document.

[wtomin]

@hadipash Hello, I think you have worked on Tensor Parallel support for diffusion models. Can you give your comments?

[wtomin]

@mxuax @ZJY0516 @SamitHuang Please leave your comments or suggested changes. Thank you very much!

[ZJY0516]

I suggest you take a look at the generated webpage — it may have some rendering issues. For example:

[hadipash]

LGTM. Very clear docs 👍