Skip to content

knowledge(vllm-omni): 1 change(s) from refine #355

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
tzhouam merged 1 commit into from
Oct 12, 2026
Merged

knowledge(vllm-omni): 1 change(s) from refine #355

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 11 additions & 2 deletions knowledge/repos/vllm-omni/models/seedvr2/rules.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,10 +1,10 @@
---
title: "SeedVR2 restoration 规则"
created: 2026-10-06
updated: 2026-10-06
updated: 2026-10-12
type: rule
tags: [vllm-omni, models, diffusion]
sources: ["PR #8102"]
sources: ["PR #8102", "PR #8767"]
confidence: high
---

Expand Down Expand Up @@ -96,3 +96,12 @@ confidence: high

公开 route 与 artifact readiness 见 [Serving 请求规则](../../components/serving/rules.md)；
分布式执行与 VAE ownership 见 [Diffusion 规则](../../components/diffusion/rules.md)。

## VLLM-OMNI-PR8767-SEEDVR2-RMSNORM-FP32 — SeedVR2 RMSNorm 的低精度统计必须在 FP32 中累计

@chatgpt-codex-connector chatgpt-codex-connector Bot Oct 12, 2026

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Route the new RMSNorm rule from the owner entrypoints

Add this nadit.py/RMSNorm trigger to both the page's Direct 代码快速入口 table and models/seedvr2/_index.md. As committed, those entrypoints only advertise clip-budget and long-video concerns, so a review starting from the bounded Direct quick map or owner index can miss the new active rule entirely, defeating its purpose for future numerical changes.

AGENTS.md reference: AGENTS.md:L23-L24

Useful? React with 👍 / 👎.


- 触发：修改 `vllm_omni/diffusion/models/seedvr2/nadit.py` 中 `RMSNorm` 的 variance 计算、dtype cast 或 affine 乘法。
- 强制：FP16/BF16 输入必须先提升到 FP32，再计算 `pow(2).mean(-1)` 与归一化除法，结果 cast 回输入 dtype 后才执行原有 affine 乘法（权重跟随输出 dtype）；FP32 输入保持原实现。该修复不要求 checkpoint 转换。
- 禁止：在输入 dtype 中对 FP16 激活求平方均值——有限值 256 平方即上溢为 inf，整个 token 除以无穷 RMS 后被静默置零；把输出-only NaN/Inf 检查当作数值验收（输出仍有限，检测不到这种 token 级置零损坏）；把单次视觉 A/B 外推为解释所有 SeedVR2 artifact 的通用结论。
- 验收：CPU 回归以 float64 参考对照，覆盖 outlier（如 256/300/60000）、全零、极小值（如 0.0001）、FP16/BF16/FP32 与 affine 开/关，断言输出 dtype、全部有限且与参考一致。^[PR #8767]

<!-- kb:rule status=active since=v0.30.0 -->
Loading