Update dsr1-fp8-mi300x-sglang SGLang image to v0.5.11-rocm700-mi30x#1348
Update dsr1-fp8-mi300x-sglang SGLang image to v0.5.11-rocm700-mi30x#1348Klaud-Cold wants to merge 1 commit into
Conversation
…n\nRef #1154\n\nCo-authored-by: Klaud Cold <Klaud-Cold@users.noreply.github.com>
|
Thanks for the contribution! For vLLM & SGLang, please ensure that your recipes is similar to the official vLLM recipes and/or the SGLang cookbook If it is not, please create a PR first before we can merge your single node PR into the master branch. Let's ensure that the documentation is first class such that the entire ML community can benefit from your hard work! Thank you PR authors are responsible for ensuring that after merging, all GitHub Action jobs fully pass. A lot of the time, failures are just flakes and simply re-running the failed jobs will fix it. If re-running failed jobs is attempted, PR authors are responsible for ensuring it passes. See GitHub's docs on re-running failed jobs: https://docs.github.com/en/actions/how-tos/manage-workflow-runs/re-run-workflows-and-jobs#re-running-failed-jobs-in-a-workflow As a rule of thumb, generally, PR authors should request a review & get a PR approval from the respective companies' CODEOWNERS before requesting a review from core maintainers. If additional help is needed, PR authors can reach out to core maintainers over Slack. |
1 similar comment
|
Thanks for the contribution! For vLLM & SGLang, please ensure that your recipes is similar to the official vLLM recipes and/or the SGLang cookbook If it is not, please create a PR first before we can merge your single node PR into the master branch. Let's ensure that the documentation is first class such that the entire ML community can benefit from your hard work! Thank you PR authors are responsible for ensuring that after merging, all GitHub Action jobs fully pass. A lot of the time, failures are just flakes and simply re-running the failed jobs will fix it. If re-running failed jobs is attempted, PR authors are responsible for ensuring it passes. See GitHub's docs on re-running failed jobs: https://docs.github.com/en/actions/how-tos/manage-workflow-runs/re-run-workflows-and-jobs#re-running-failed-jobs-in-a-workflow As a rule of thumb, generally, PR authors should request a review & get a PR approval from the respective companies' CODEOWNERS before requesting a review from core maintainers. If additional help is needed, PR authors can reach out to core maintainers over Slack. |
| - config-keys: | ||
| - dsr1-fp8-mi300x-sglang | ||
| description: | ||
| - "Update SGLang image from v0.5.9-rocm700-mi30x to v0.5.11-rocm700-mi30x" | ||
| pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/XXX |
There was a problem hiding this comment.
🟡 The new perf-changelog entry uses a placeholder PR link https://github.com/SemiAnalysisAI/InferenceX/pull/XXX instead of the actual PR number (#1348). Every other entry in this file uses a real numeric PR ID, so this leaves a permanently broken (404) cross-reference in changelog history once merged. Please replace XXX with 1348 before merging.
Extended reasoning...
What the bug is
The diff appends a new entry to perf-changelog.yaml for dsr1-fp8-mi300x-sglang:
- config-keys:
- dsr1-fp8-mi300x-sglang
description:
- "Update SGLang image from v0.5.9-rocm700-mi30x to v0.5.11-rocm700-mi30x"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/XXXThe pr-link field contains the literal placeholder XXX rather than a numeric PR id. This appears to be a template that the author (or a generator) intended to fill in but forgot to substitute.
Why this is wrong
Every other entry in perf-changelog.yaml uses a real numeric GitHub PR id. The five entries immediately preceding this one use /pull/1310, /pull/1308, /pull/1305, /pull/1304, and /pull/1303 respectively, establishing a strong and unambiguous convention. This PR is #1348 (per the PR metadata), so the link should be https://github.com/SemiAnalysisAI/InferenceX/pull/1348.
Impact
Functionally this is benign — the YAML still parses and any consumer that treats pr-link as an opaque string will not fail. The harm is purely to the changelog as documentation: clicking the link will resolve to GitHub's /pull/XXX path, which is not a valid PR number and returns a 404. Because the changelog is append-only history, once this merges the broken link is permanent unless a follow-up commit corrects it.
How to fix
Replace XXX with 1348 on line 2351:
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1348Step-by-step proof
- Look at
perf-changelog.yamlline 2351 in the diff:pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/XXX. - Look at the PR metadata: this PR is number 1348.
- Look at the existing convention in the file — the preceding entry (line 2345 in the new file) reads
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1310, a real numeric id pointing at PR [DEBUG] Add Qwen3.5-397B-A17B FP8 MI355X ATOM perf-changelog entries #1310. - Construct the URL the changelog will produce after merge:
https://github.com/SemiAnalysisAI/InferenceX/pull/XXX. GitHub's pull-request route requires an integer;XXXis not an integer, so the link returns 404. - Therefore the changelog as written contains a permanently broken cross-reference to this PR.
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=25763467998 |
1 similar comment
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=25763467998 |
Summary
dsr1-fp8-mi300x-sglangimage fromlmsysorg/sglang:v0.5.9-rocm700-mi30xtolmsysorg/sglang:v0.5.11-rocm700-mi30xRef #1154
Generated with Claude Code