Skip to content

Docs: refresh README quickstart, architecture and release guidance - #533

Merged
Dong1017 merged 9 commits into
mainfrom
codex/refresh-readme-531
Nov 21, 2026
Merged

Dong1017 merged 9 commits into
mainfrom
codex/refresh-readme-531

Conversation

@Dong1017

@Dong1017 Dong1017 commented Nov 20, 2026 •

Copy link
Copy Markdown
Collaborator

Closes #531

What changed

  • First screen for busy readers. Lead with one sentence on what InferRSI is and who it is for, a copy-paste quickstart that goes from a fresh clone to a runnable CLI, and a real run-review dashboard screenshot — so a developer can judge the tool and try it within one screen.
  • Selected results. Replace the old results dump with five compact tables (onboarding, text serving, image and video, speech and multimodal, standalone engines) that keep measured gains of at least 20%; add docs/optimization-results.md for the full curated record. One 🔥 per table marks the largest comparable gain, and ✅ appears only where the linked evidence includes output-quality checks. The DeepSeek V4.1 Flash W4A8 row states its scope explicitly: throughput and latency were rechecked at each scenario's latency limit, while long-output accuracy remains unverified, so it carries no quality-pass mark.
  • Architecture and compatibility (F1, F3, F4, S1, D1). Link the offline controller to the Apps workflow, Benchmark acceptance rules and Harness candidate selection. Refresh the diagrams shared by the README and architecture guide, correct the host inspection importer and repository-pack path, and point Python callers to the import migration list. Rendered architecture section.
  • Getting started (M1–M4). Run the CPU target check and list packaged tasks before searching. Document the five backend choices, the opt-in CANNbot checkout and interpreter, the separate MindIE reader and how to revoke either project grant. Account for the knowledge reorganization that landed after the issue audit, including the availability of skills enter after v0.2.6 and the current knowledge entry points.
  • Review follow-ups. Use inferrsi dashboard --input . with the default dashboard.html so the dashboard never overwrites the report.html written by run; keep one 🔥 per performance table; and align every ✅ mark with the scope of its cited evidence.

Verification

  • Merged the latest main into this branch and resolved the one conflict in docs/architecture.md, keeping the current apps/integrations/cannbot/ layout. All eight files in the diff are documentation or diagram assets; no code changes.
  • Built and installed from this checkout on Linux with Python 3.12.3. Ran inferrsi --help, benchmarks list (returns moe-forward-npu, qwen21-denoise-npu, qwen3-tts-npu), the CPU doctor check against a vLLM-Omni checkout, the README find query and both context exports (five records each). Confirmed the cannbot, mindie and skills enter command surfaces.
  • Checked every relative README link on the merged tree: 47 unique targets resolve with none missing, including 10 cross-document anchors, plus 41 external HTTP targets. The quickstart block shows the real benchmarks list output.
  • Rechecked the 11 linked vLLM-Omni PRs on 2026-11-20: #8739, #8729, #10184, #9218, #9188, #9827 and #8112 merged; #9698, #9745 and #9283 open drafts; #10243 open. docs/optimization-results.md carries those labels.
  • Re-audited every ✅ in the results tables against its cited evidence. Each remaining ✅ names output-level checks (quality comparisons, pixel-identical or numerical outputs, bit-identical audio, contract checks); the DeepSeek throughput row now states its limited scope instead of claiming a quality pass.
  • The refreshed SVG diagrams are valid XML, and every path they show exists under src/inferrsi/ on main.

The issue owner is submitting this revision for review and closure on merge. Verification covered documentation, CLI behavior and upstream PR metadata; performance and quality figures retain their original campaign evidence. GUI plugin installation was outside this run.

Align the shared diagrams, paths and import guidance with the current package tree. Document the checkout-scoped knowledge grants and distinguish the released CLI from task entry on main.

Agent-assisted-by: Codex (author)
Signed-off-by: Dong1017 <55723162+Dong1017@users.noreply.github.com>
@Dong1017 Dong1017 added the kind:docs Documentation and template delivery label Nov 20, 2026
zengchuang-hw
zengchuang-hw previously approved these changes Nov 20, 2026

@zengchuang-hw zengchuang-hw left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

PASS

初次独立检视 1b826b5f → 23be5b72,逐项看了六处文档/图示变更与 #531 的修订:三个上游 Draft 标注仍保留,#10184 更新为已合入;Find 只说明当前注册的 Cannbot,MindIE 的 CLI/reader 单列。旧 import 路径的兼容边界、图示资源、当前架构链接、CLI 参数和授权路径也与源码对齐。无批准前必修项。

非阻塞:可把 v0.2.6 标为 preview,和发布页措辞一致。已静态核对源码、文档链接及 11 个上游 PR 状态;未执行 CLI、渲染图示或查询 CI。

cursor[bot]
cursor Bot previously approved these changes Nov 20, 2026

@cursor cursor Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

LGTM

PR review verdict — PASS
Reviewed base 1b826b5f9666a8b734000e1eb2990e582cdef876 and head 23be5b7267ccb3d45e05d623a830fb13912b5a57.

Independent third-party technical verify (cursor[bot]; collaborator permission none — not a write+ eligible Approver under docs/agent-collaboration.md; does not complete dual tip / merge-ready).

Verified against governing #531 slice and tip tree:

  • README Results table: #10184 marked merged; #9698 / #9745 remain draft; #10243 open — matches the PR claim of rechecked upstream statuses.
  • Architecture / docs/contracts/current-design.md / contributing: host inspection importer and harness/data/repo_packs path align with src/inferrsi/apps/host/inspection.
  • SVG assets remain well-formed; Ready tip green (full-check 3.10/3.11, declarations, installed-package, dashboard-browser, pr-references).

Peer tip Approve: zengchuang-hw. Author is Dong1017 → Dong ineligible to Approve; this is cursor first formal tip vote as independent verify.

Merge still needs write+ independent Approves / steward integration — not claimed here.

amy-why-3459
amy-why-3459 previously approved these changes Nov 20, 2026

@amy-why-3459 amy-why-3459 left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

LGTM

PR review verdict — PASS
Reviewed base 1b826b5f9666a8b734000e1eb2990e582cdef876 and head 23be5b7267ccb3d45e05d623a830fb13912b5a57.

Docs-only refresh for #531. Verified:

  • The v0.2.6 release exists (published 2026-11-20) and the README highlights match its notes (paired knowledge-on/off reports and service-reuse evaluation).
  • Upstream statuses rechecked: vLLM-Omni #10184 is merged; #9698 and #9745 remain drafts; #10243 is open, matching the 2026-11-20 table.
  • apps/workflows/improvement.py, apps/host/inspection.py, and harness/data/repo_packs path updates match this checkout. skills enter is correctly described as introduced after v0.2.6.

Accuracy and performance comparisons are not applicable (documentation only). No outstanding CHANGES_REQUESTED on this head.

Non-blocking: the v0.2.6 bullet could note that the GitHub release is a preview, consistent with the release page.

Copy link
Copy Markdown
Collaborator

Thanks for the first-screen quickstart and the merge onto current main.

Reviewed 23be5b7267ccb3d45e05d623a830fb13912b5a57 → e520d79f385d266ee13e04283b0e9dadf725f6e4 (PR base 6bdc86ee339c3007f15833434065adf893bff953).

The merge conflict in docs/architecture.md correctly keeps the post-#528 apps/integrations/cannbot/ layout. The new clone-and-inferrsi benchmarks list block matches tests/test_engineering_benchmarks.py (["moe-forward-npu", "qwen21-denoise-npu", "qwen3-tts-npu"]), and docs/assets/dashboard-knowledge-default-run-review.png plus the Codex plugin #install-from-the-release-zip anchor are present.

One first-screen wording issue to fix before I can approve:

  • In README.md L26, “Every recorded run also produces a local evidence dashboard” overstates what the tool does. docs/dashboard.md documents the dashboard as a view generated by inferrsi dashboard --input <RUN_OR_CAMPAIGN_DIRECTORY>, not as an artifact written automatically by every recorded run. Please rephrase this (and keep the screenshot) so a first-time reader is not led to expect a dashboard file from the run itself — for example, that a recorded run can be reviewed with inferrsi dashboard --input ….

Accuracy and performance comparisons are not applicable (documentation only). No outstanding CHANGES_REQUESTED on this head.

@zengchuang-hw zengchuang-hw left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

NO PASS

增量检视 23be5b7267ccb3d45e05d623a830fb13912b5a57 → e520d79f385d266ee13e04283b0e9dadf725f6e4,本轮 base 6bdc86ee,承接上轮检视。之前对 Draft、MindIE 和 CANNBot Find 的区分仍成立;v0.2.6 能力表述和 quickstart 入口也已核对。

批准前必做:README 首屏新加的 “Every recorded run also produces a local evidence dashboard” 把独立的 dashboard 说成了 run 的默认产物。实际 run 返回的是 report.html(src/inferrsi/apps/cli/__init__.py:772–783),截图所示的 dashboard 由另一个命令打开或导出(同文件 :614–628)。建议改成 “Open a recorded run in the local evidence dashboard with inferrsi dashboard --input <run-directory>.” 将自动生成的报告与 dashboard 区分。

已核对六个文档/图示变更及相关 CLI 源码;未执行 CLI 或测试。

Use aligned result tables with explicit baselines, preserve validation limits, and link the benchmark and historical records. Refresh the README banner and collaboration guidance.

Signed-off-by: Dong1017 <55723162+Dong1017@users.noreply.github.com>
Agent-assisted-by: Codex (author)
Agent-assisted-by: Codex (author)
Signed-off-by: Dong1017 <55723162+Dong1017@users.noreply.github.com>

Copy link
Copy Markdown
Collaborator

Reviewed e520d79f385d266ee13e04283b0e9dadf725f6e4 → c242ac5e7f7b3db09201a82021b58937049de4b2 (PR base 6bdc86ee339c3007f15833434065adf893bff953).

The previous first-screen wording issue is resolved. This README now tells readers to generate a dashboard from a recorded run directory with inferrsi dashboard --input . --output report.html, which matches docs/dashboard.md#generate-and-open. Non-blocking: the documented default output is dashboard.html; naming the example report.html may again blur the dashboard with the HTML report written by run.

This increment is larger than that wording fix. e55eacfb / c242ac5e redesign the README, add docs/optimization-results.md, and introduce newly curated “≥20%” tables plus several results that were not in the previous README (DeepSeek V4.1 Flash, MiniCPM #431, Drama #132, Qwen3-TTS #9068, and the #520 950PR row). The new doc links and anchors I checked resolve.

Please address the following before I can approve:

  1. Inconsistent LFM3 percentage. In both README.md (Selected Results) and docs/optimization-results.md, LFM3-Omni-2B is written as 0.285 → 0.593 requests/s (+108.4%). (0.593 − 0.285) / 0.285 = 108.1%. Please correct the percentage or the endpoints so they agree, and cite any unrounded source values if you keep 108.4%.

  2. Typo in docs/optimization-results.md. The DeepSeek W4A8 prefill metric is labelled “Input throughout the run”. If this means throughput, please fix the wording.

  3. “Historical records” dropped previously listed campaigns. docs/optimization-results.md opens with “The tables show measured improvements of at least 20%,” and the moved vLLM-Omni table no longer includes the previously listed sub-20% rows:

    • MiniMax-H3 Ref2VA on 4 logical Ascend 910C devices, 291.25 → 286.67 s (−1.57%), campaign INCONCLUSIVE, vLLM-Omni #9745 (draft)
    • Qwen3-Omni cuDNN pilot, +7.79% throughput with a TTFA regression, #9283 (draft)
    • LingBot-World KV gather, 108.29 → 103.89 s (−4.06%), #8112 (merged)

    A 20% highlight in the README is fine, but the README links this page as “Benchmark details and historical records.” Those historical rows (and their PR statuses) should stay on the details page, or the intro should not present the page as the complete record. Docs: refresh README items that no longer match the repository (2026-11-20 audit) #531 F2 was specifically about keeping the linked upstream PR labels accurate, including the drafts.

  4. New DeepSeek onboarding numbers. The README Model onboarding row cites a six-run baseline of 39.28 / 44.07 / 56.86 ms/token at C1 / C4 / C16, linked to a #:~:text= fragment for 9be3c22… on acceptance: DeepSeek V4.1 Flash 910C 8K/1K TPOT ≤25 ms/token at C1/C4/C16 #353. Those figures were not in the previous README, they are not in the acceptance: DeepSeek V4.1 Flash 910C 8K/1K TPOT ≤25 ms/token at C1/C4/C16 #353 issue body, and they differ from the historical W8A8 pilot numbers now in docs/optimization-results.md (which is appropriate if they are the acceptance: DeepSeek V4.1 Flash 910C 8K/1K TPOT ≤25 ms/token at C1/C4/C16 #353 follow-up baseline). Please point to the exact acceptance: DeepSeek V4.1 Flash 910C 8K/1K TPOT ≤25 ms/token at C1/C4/C16 #353 comment or report that establishes these three values as the six-run baseline so they can be checked.

Non-blocking:

  • The PR description still describes the earlier e520d79f README (News / v0.2.6, the 11-PR table, architecture diagrams, first-screen dashboard screenshot). Please update it to match this revision.
  • In the h3-metal details table, original-H3 vs h3.c cold changed from 1.25× to — (the cold time reduction is about 19.7%, just under 20%). A one-line note would help readers who expect the previous figure.

Accuracy and performance comparisons are not applicable as new experiment evidence (documentation only); the points above are about how existing numbers are transcribed and curated. No outstanding CHANGES_REQUESTED on this head.

@zengchuang-hw zengchuang-hw left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

NO PASS

增量检视 e520d79f385d266ee13e04283b0e9dadf725f6e4 → c242ac5e7f7b3db09201a82021b58937049de4b2,承接上轮检视。已经明确 dashboard 是单独运行的命令,上一轮措辞问题已修复。

批准前必做:README 的新示例从已有运行目录执行 inferrsi dashboard --input . --output report.html。但普通 inferrsi run 已生成 report.html,而 apps/views/dashboard.py:383–411 的导出会拒绝覆盖非 dashboard HTML。请用 inferrsi dashboard --input . 默认输出 dashboard.html,或显式指定 --output dashboard.html,保留原报告。

已静态核对本轮 README/结果文档/图示差异、CLI 实际输出及 dashboard 覆盖保护;未运行 CLI 或测试。

- Add emoji to section headings and keep every previous anchor id so
  existing links into the README still resolve.
- Merge the Baseline column into the measured-change column, taking the
  result tables from five columns to four for narrow screens, and shorten
  the validation cells.
- LFM3-Omni-2B: report +108.1%, which is what the listed 0.285 -> 0.593
  req/s endpoints give; the details page notes the upstream +108.4% comes
  from unrounded means.
- DeepSeek W4A8 prefill: "Input throughput" instead of "Input throughout
  the run".
- Restore the three below-20% vLLM-Omni campaigns to the details page in
  their own section, with their draft and merged labels intact, and stop
  describing the page as only showing 20%+ tables.
- Point the six-run DeepSeek baseline at the comment that states
  39.28 / 44.07 / 56.86 ms/token.
- Dashboard example: use `inferrsi dashboard --input .` and its
  dashboard.html default, since `run` already writes report.html and the
  exporter refuses to overwrite a document it did not generate.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Dong1017 <55723162+Dong1017@users.noreply.github.com>

Copy link
Copy Markdown
Collaborator

PR review verdict — DO NOT PASS

Reviewed c242ac5e7f7b3db09201a82021b58937049de4b2 → 43d313545eb40a26d891a0720054963c78159784 (PR base 6bdc86ee339c3007f15833434065adf893bff953).

The four items from my previous comment are addressed:

  1. LFM3 is now +108.1% in both README.md and docs/optimization-results.md, which matches (0.593 − 0.285) / 0.285.
  2. The W4A8 prefill metric is now “Input throughput”.
  3. docs/optimization-results.md now keeps the three sub-20% historical rows (#9745 draft, #9283 draft, #8112 merged) and describes the page as the full curated record.
  4. The DeepSeek onboarding figures now link to #353 comment 6641425447, where the six-run median TPOT values 39.282… / 44.072… / 56.862… round to 39.28 / 44.07 / 56.86 ms/token at C1 / C4 / C16.

The dashboard example now uses inferrsi dashboard --input . and dashboard.html, matching docs/dashboard.md#generate-and-open. The h3-metal cold-time note is also in place.

Please fix the following before I can approve:

  1. Misplaced LFM3 percentage note. In docs/optimization-results.md L104–106, the sentence “The LFM3-Omni-2B change is computed from the rounded means above; upstream quotes +108.4% from its unrounded values.” is appended to the glossary after the 950PR step-reduction table, whose “means above” are 5.04× / 7.37× / 4.74× / 4.34×, not 0.285 → 0.593. Please move this note next to the LFM3 row (L61).

  2. 🔥 legend does not match the marks. README.md L92 says “🔥 marks the largest gain in its table,” but the Image and video table has three 🔥 marks (L119–121: −73.45%, −76.3%, −60.16%) and the Speech table has two (L132 +108.1% and L136 +158%). Please either keep a single 🔥 in each table (the largest comparable gain) or rephrase the legend so it matches the current marks.

Non-blocking: the PR description still describes the earlier e520d79f README (News / v0.2.6, the 11-PR table, architecture diagrams, first-screen dashboard screenshot) and still links the architecture section at that commit. Please update it to match this revision.

Accuracy and performance comparisons are not applicable as new experiment evidence (documentation only). No outstanding CHANGES_REQUESTED on this head.

- README: the legend says the fire mark is the largest gain in its table,
  so only the 4.22x row, the +158% row and the +381% row keep it.
- optimization-results: the sentence about the LFM3 percentage cited
  "the rounded means above", which sat under the 950PR step-reduction
  table. Move it into the LFM3 row itself.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Dong1017 <55723162+Dong1017@users.noreply.github.com>
amy-why-3459
amy-why-3459 previously approved these changes Nov 20, 2026

@amy-why-3459 amy-why-3459 left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

LGTM

PR review verdict — PASS

Reviewed 43d313545eb40a26d891a0720054963c78159784 → a777b007bbd4e4c9074c46f93c4091222c2e4c00 (PR base 6bdc86ee339c3007f15833434065adf893bff953).

Both items from my previous comment are resolved:

  1. The LFM3-Omni-2B rounding note (“Computed from the rounded means; upstream quotes +108.4% from its unrounded values.”) now sits in the LFM3 row of docs/optimization-results.md (L61), and the glossary after the 950PR step table only defines TTFA / RTF / WER.
  2. README.md now has exactly one 🔥 per performance table that carries one, each on its largest gain: Image and video −76.3% (4.22×, larger than 3.77× / −73.45%), Speech +158% (Qwen3-TTS 1.7B, above +108.1% / +100.6%), and Standalone engines +381%. This matches the legend at L90.

The increment is a single docs-only commit (2 files, +5/−7); knowledge-model-policy is success on this head, and there are no open CHANGES_REQUESTED or inline threads on it. Accuracy and performance comparisons are not applicable (documentation only).

Non-blocking (unchanged from last time): please refresh the PR description, which still describes the earlier e520d79f README.

@zengchuang-hw zengchuang-hw left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

NO PASS

增量检视 c242ac5e7f7b3db09201a82021b58937049de4b2 → a777b007bbd4e4c9074c46f93c4091222c2e4c00,承接上轮。Dashboard 示例已改为默认写 dashboard.html,不再撞上原报告。

批准前必做:README 新图例把 ✅ 定义为通过质量检查,却给 DeepSeek V4.1 Flash 三组 W4A8 吞吐复测加了同一标记。已核对吞吐数字,但 纯 P 原始证据明确保留长输出精度未验证,延迟门限内的吞吐复测不能作为整行质量通过。请移除这个 ✅,或按场景注明实际覆盖的质量范围。吞吐值不需要改。

已对照本轮 README、结果文档和原始证据;未运行 CLI 或复测。

@Dong1017

Copy link
Copy Markdown
Collaborator Author

Thanks @zengchuang-hw. Fixed in 13c1d4a.

The DeepSeek V4.1 Flash W4A8 row under Text serving no longer carries ✅. Its validation cell now reads: "Throughput and latency rechecked at each scenario's latency limit; decode rechecked once. Long-output accuracy not verified." The three throughput figures, their baselines and the +54.45% / +36.12% / +20.80% conclusion are untouched.

Your reading of the source matches mine: case19-pure-p.md still lists cross-restart, long-run and long-output accuracy as pending verification, so a latency-bounded throughput recheck cannot support the legend's "quality checks passed" meaning.

I also re-audited every other ✅ in the README results tables. Each remaining ✅ in the performance tables names its output-level check (14/14 quality comparisons, 9 and 44 pixel-identical outputs, 30/30 numerical checks, bit-identical audio, audio and contract checks), so none of them rests on throughput-only evidence. The onboarding LFM3-Omni-2B ✅ covers the two real-weight serving tests named in the same cell.

The PR description now describes head 13c1d4a, including the architecture link, and the branch was pushed without force. No other rows or numbers changed.

@zengchuang-hw @amy-why-3459 could you take another look once CI is green?

zengchuang-hw
zengchuang-hw previously approved these changes Nov 20, 2026

@zengchuang-hw zengchuang-hw left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

PASS

增量检视 a777b007bbd4e4c9074c46f93c4091222c2e4c00 → 13c1d4a37c684fa5c4b26f25442715228c6a3822,承接上轮检视。本轮只改 README 的 DeepSeek 行:移除质量通过标记,明确吞吐和延迟复核的范围以及长输出精度未验证。对照纯 P 实验原文,该限定与证据一致,原有吞吐数字未改。Dashboard 示例仍默认写入 dashboard.html,不会与原报告冲突。

本轮没有批准前必做项。已静态核对精确差异、README、证据及 Dashboard 输出路径;未运行 CLI、复测或检查 CI。

amy-why-3459
amy-why-3459 previously approved these changes Nov 20, 2026

@amy-why-3459 amy-why-3459 left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

LGTM

PR review verdict — PASS

Reviewed a777b007bbd4e4c9074c46f93c4091222c2e4c00 → 13c1d4a37c684fa5c4b26f25442715228c6a3822 (PR base 6bdc86ee339c3007f15833434065adf893bff953).

The only change is the DeepSeek V4.1 Flash row under Text serving in README.md (L111): the ✅ is removed and the cell now scopes the evidence to throughput and latency rechecks, with “Long-output accuracy not verified.” This matches case19-pure-p.md, which says cross-restart, long-run and long-output accuracy are still pending, and it keeps the ✅ legend consistent. The +54.45% / +36.12% / +20.80% figures and baselines are unchanged.

The PR description now describes this head. knowledge-model-policy is success on 13c1d4a3, and no CHANGES_REQUESTED or other blocking review is open on this head. Accuracy and performance comparisons are not applicable (documentation only).

@Dong1017
Dong1017 dismissed stale reviews from amy-why-3459 and zengchuang-hw via e1d9065 November 20, 2026 23:46

@zengchuang-hw zengchuang-hw left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

PASS

增量检视 13c1d4a37c684fa5c4b26f25442715228c6a3822 → e1d9065ff9881c43cfb478fa0a8aa235a018d4a1,承接上轮检视。本轮同步了已落地的 #535;以 fb343547f5c72155a8554038773866327da8d446 为基线,实际 PR 仍是 8 个文档和图示文件。合并保留了 Anthropic 集成与 artifact helpers 的新归属、README 对 DeepSeek 长输出精度的限定,以及默认输出 dashboard.html 的示例。

本轮没有批准前必做项。非阻塞:主干继承的架构表仍有一行旧 Model 职责,与下文的包迁移说明略重复,可以另行归并到 Apps。已静态核对精确 head、提交范围、实际差异和相关路径;未运行 CLI、实验或检查 CI。

@amy-why-3459 amy-why-3459 left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

LGTM

PR review verdict — PASS

Reviewed 13c1d4a37c684fa5c4b26f25442715228c6a3822 → e1d9065ff9881c43cfb478fa0a8aa235a018d4a1 (new PR base fb343547f5c72155a8554038773866327da8d446, i.e. main after #535).

The increment is only the merge of main (#535). Against the new base the PR still changes the same 8 documentation/diagram files; the patches for README.md, docs/optimization-results.md, docs/contracts/current-design.md, docs/contributing.md and the three SVGs are byte-identical to the previously approved 13c1d4a3 diff. The only resolved file is docs/architecture.md: the merge drops this PR's earlier Package table (which still listed Model/Infra packages) and keeps main's post-#535 table and text, while keeping this PR's intro, migration-list/import-gate links and the harness/data/repo_packs fix. Relative links and anchors in README.md and docs/architecture.md resolve on this head, and none of the PR's files reference the removed inferrsi.infra / inferrsi.model paths.

Accuracy and performance comparisons are not applicable (documentation only). No CHANGES_REQUESTED or other blocking review is open on this head; full-check was still running when I looked.

Non-blocking (inherited from main, not introduced here): docs/architecture.md L98 still has a Model row although #498/#535 removed that package, and its provider adapters are now listed under Apps; docs/contributing.md L103 still says "Harness/Model/Infra plane". Both could be folded into a follow-up.

@Dong1017
Dong1017 merged commit 6ed42af into main Nov 21, 2026
11 checks passed
@Dong1017
Dong1017 deleted the codex/refresh-readme-531 branch November 21, 2026 02:03
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

kind:docs Documentation and template delivery

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Docs: refresh README items that no longer match the repository (2026-11-20 audit)

3 participants