Repository navigation
Qualify ATOM cache reuse against profiling requests / 核对 ATOM 缓存复用的 profiling 请求范围 - #4405
Open
chunfangamd wants to merge 8 commits into
Open
chunfangamd wants to merge 8 commits into
chunfangamd wants to merge 8 commits into
Conversation
保留 ATOM 窗口与请求范围合同，复用 counter 修复并隔离诊断数据。
保留缺少声明端点观测时的旧资格状态。
Contributor
|
Generally most InferenceX submissions don't need to add changes to the infx package. On the off chance they do, generally the diff is small. Please prompt your agent to clean your PR to keep your diff to only necessary & minimal changes too. We have an increased code quality bar for changes to the infx package. 中文大多数 InferenceX 提交通常不需要修改 infx 包。即使确实需要，改动通常也很小。请让你的 agent 清理 PR，使改动只包含必要且最小的内容。infx 包的改动有更高的代码质量要求。 |
Contributor
|
Claude finished @chunfangamd's task in 39s —— View job LGTM - no blocking issues found I couldn't run the test suite here because |
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
合入独立 counter 修复，保留现有 ATOM 范围检查及公开历史。 # Conflicts: # inferencex-e2e/infx/results/agentic/server_metrics.py
将 null labels 作为无标签域处理，继续拒绝非法结构与重复别名。令成功范围的 prompt_total 与缓存来源分解使用相同 full 计数，并删除新增模块概述注释。
中文：删除旧窗口恢复与冗余诊断，保留现代 ATOM 请求守恒与最小共享资格合同。
中文：现代 P/D 和 unified 正例包含不完整传输诊断，资格与核心命中率保持有效。
确认 CPU offload 关闭时允许已知零值与缺失计数域共存，同时拒绝非零、非法或额外计数域；启用 offload 时继续要求完整覆盖。
中文：同步最新 ATOM 标签域计数修复与 main，保留现代采集器资格判断及上下文传递。
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Correct counters still need a profiling scope before they can support measured pricing. This layer qualifies ATOM exports using the verified AgentX 1.0.12 profiling contract, matching counter domains and successful client/server request and prompt-token totals. It emits the small validity core consumed by app #1377.
A confirmed disabled CPU tier accepts both absent counters and a mixture of valid zero/absent domains. Invalid present counters and unexpected domains still fail qualification. Transfer diagnostics do not decide cache validity.
CPU replay through the actual producer CLI preserves 60% GPU / 10% CPU / 70% overall for the qualified saved collector export and rejects the original server29/client30 scope mismatch. This is saved/simulated collector evidence, not a live GPU P/D/offload acceptance run. Legacy telemetry remains visible but ineligible for measured pricing; historical recovery is outside this PR.
Base: #4397. Next: #4163. The app's default remains legacy; merging this producer does not activate consumer enforcement.
中文
计数正确还需要证明 profiling 范围。本层依据已核实的 AgentX 1.0.12 导出合同、匹配计数域及成功请求/token 守恒，为 ATOM 生成 app #1377 消费的小型 validity core。确认关闭 CPU tier 后，完整缺失与合法零/缺失混合均可接受；坏值、配置矛盾及额外域仍拒绝，传输诊断不参与资格判定。
实际 producer CLI 重放保留合法输入的 60%／10%／70%，并拒绝原始 server29/client30 范围不一致。这是保存的模拟 collector 证据，不代表真实 GPU P/D/offload 验收。旧数据仍有遥测，但不授予实测计价资格；不恢复历史窗口。依赖 #4397，后续为 #4163。app 默认仍 legacy，合并 producer 不会启用消费端校验。
AI model disclosure
GPT-6 (Codex), including delegated implementation and independent review agents, prepared and checked these revisions. A finer runtime model identifier is not exposed. Earlier producer/app work was prepared by an agent reporting Claude Opus 5.5; that historical runtime identifier could not be independently verified.