Repository navigation
knowledge(vllm-omni): 1 change(s) from refine - #340
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 97b96fe546
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
|
||
| ## VLLM-OMNI-PR8510-MHC-FP32-MUSA — MAGI-2 mHC 收缩的 MUSA 形态必须保留 einsum 回退、误差界与 stream-major 布局 | ||
|
|
||
| - 触发：修改 `vllm_omni/diffusion/models/magi2/layers.py` 中 `MHCHandler` 的 `apply_pre`、stream mix（`hyper_connect`）、`compute_logits`、`_stream_major_normed` 或 `__init__` 里的平台分派；或改动 mHC stream contraction 的精度、dtype 与内存布局。 |
There was a problem hiding this comment.
This contract applies exclusively to vllm_omni/diffusion/models/magi2/layers.py, while the repository already has knowledge/repos/vllm-omni/models/magi2/rules.md; placing it in the shared diffusion page means model-owner routing can omit it and incorrectly presents a model-specific optimization as a cross-model invariant. Move the rule and its source attribution to the MAGI-2 owner page and update that owner's _index.md.
AGENTS.md reference: AGENTS.md:L12-L17
Useful? React with 👍 / 👎.
| `fullgraph=True`。验证 tensor 内容变化不重编译时固定 host scalar 和 history phase， | ||
| 将空历史→满历史引发的合法 scalar specialization 与 tensor recompile 分开。 ^[PR #7648] | ||
|
|
||
| ## VLLM-OMNI-PR8510-MHC-FP32-MUSA — MAGI-2 mHC 收缩的 MUSA 形态必须保留 einsum 回退、误差界与 stream-major 布局 |
There was a problem hiding this comment.
The heading uses VLLM-OMNI-PR8510-MHC-FP32-MUSA, but this page declares that only DIFF-<number><optional-letter> identifiers are auditable, and knowledge/tools/check_review_report.py::extract_rule_ids accepts only that stable-ID form. Consequently Direct review reports cannot select or validate this rule even though the prose was added successfully; rename it to a valid owner-specific stable ID and expose it in the corresponding quick map.
Useful? React with 👍 / 👎.
Generated automatically by the Copilot knowledge service (changeset
vllm-omni-refine-b305c07b108f).repos/vllm-omni/components/diffusion/rules.mdGenerator:
zcode:GLM-5.3· judge:codex:gpt-6-sol:medium.The publisher merges it only if its local gate passes on the exact merge result.