[Model] Support Step-3.5-Flash

[Joeegin]

PR Description

*完善路由缩放参数支持
*完善Grouped TopK校验
*添加一个环境变量（KUNLUN_DISABLE_SMALL_MOE）,手动设置成1，禁用小规模Moe分支优化（否则会RuntimeError: CUDA error: unspecified launch failure）
*新增update_block_table方法

Doc：

• Add step-3.5-Flash tutorials document

---

Checklist (Required)

Before submitting this PR, please ensure that all the following items are completed:

• All code changes pass the pre-commit checks.
• Commits are signed off using git commit -s.
• The PR title is properly classified (see below).

---

PR Type

Please prefix the PR title with one or more of the following labels to help reviewers quickly understand the nature of the change:

• [Feature] – New features or enhancements (e.g. Attention, Communicator, Kernel, Worker, etc.)
• [Bugfix] – Bug fixes
• [CI/Build] – CI, build system, or infrastructure improvements
• [Doc] – Documentation updates or fixes
• [Misc] – Other changes that do not fit the above categories (use sparingly)

Note: If the PR spans multiple categories, include all relevant prefixes.

---

Detailed Checklist (Click to Expand)

Thank you for contributing to vLLM Kunlun! To help us maintain high code quality and streamline the review process, please ensure your PR meets the following requirements.

1. Code Quality

• All linting and formatting checks pass (pre-commit).
• The code is well-structured and sufficiently documented.
• The change is designed with maintainability and readability in mind.

2. Testing

• Relevant unit tests are added or updated.
• Integration tests are included when applicable.
• Existing tests continue to pass.

3. DCO Compliance

This project follows the Developer Certificate of Origin (DCO).

• All commits include a Signed-off-by: line.
• Use git commit -s to automatically add the sign-off.

4. Review Expectations

During the review process, maintainers may:

• Request code refactoring or additional tests.
• Ask for clarifications on design decisions.
• Suggest performance, stability, or maintainability improvements.

We appreciate your patience and collaboration throughout the review process!

[Joeegin]

[Copilot]

Pull request overview

Adds support and stability fixes for Step-3.5-Flash in the Kunlun backend, primarily around MoE routing/scaling and attention metadata handling.

Changes:

• Add update_block_table to KunlunAttentionMetadataBuilder and enforce contiguous prefill Q/K/V slices.
• Extend Kunlun fused MoE to support a router scaling factor and improve grouped-TopK parameter validation.
• Add KUNLUN_DISABLE_SMALL_MOE=1 escape hatch to disable the small-MoE fast path.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
vllm_kunlun/v1/attention/backends/kunlun_attn.py	Adds metadata update helper and fixes prefill tensor contiguity; introduces a new stdlib import.
vllm_kunlun/ops/fused_moe/layer.py	Plumbs routed scaling factor from the layer into Kunlun fused MoE calls.
vllm_kunlun/ops/_kunlun_ops.py	Adds router_scaling_factor, validates grouped-topk params for sigmoid routing, and adds env-var gating for small-MoE path.

---

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