Repository navigation
[VideoGenX] include video configs in local change detection and rendering / [VideoGenX] 本地变更检测与渲染纳入视频配置 - #4886
Merged
Merged
Conversation
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。 |
4 of 11 tasks
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
python -m infx.matrix.changedandinfx generate --config-keysee no video points. #2916 madevideo-generationopt-in ingenerate_config_matrix, so a key's video scenarios expand only whenscenario_typesnames them.changed.head_matrix,changed.base_matrixandgenerate_recipescall the generator with no scenario selection, so every*-videokey has zero points on both sides of the comparison and nothing to render.One instance, measured on origin/main at 04118aa: after an engine-arg edit to
h3/sglang/h100-bf16/serving-client-tp2.yaml,changed --base origin/main --jsonprintschanged: []across 184 keys, whiletest-config --scenario-type video-generationmoves theg2fingerprint fromf152db5d9170toc1f805258915.infx generate --config-key h3-bf16-h100-sglang-videoexits 2 withhas no srt-slurm points. The PR template checklist and the review bot rely onchanged, so a video recipe change looks like it needs noperf-changelog.yamlentry and no video sweep runs.Changes:
infx.matrix.generate.SCENARIO_TYPESis the one list of scenario types a changelog entry can select. The--scenario-typechoices andplan.pyread it from there;plan.DEFAULT_SCENARIO_TYPESis unchanged.changed.config_pointsgenerates every scenario type of one key, andhead_matrixuses it, so the working tree's video points carry recipe fingerprints.changed.base_matrixkeeps the one unscopedtest-configrequest per key for the base revision's LLM scenarios and adds one--scenario-type video-generationrequest for each key that declares a video scenario, through_base_video_points. The base generator exits when a video recipe cannot be bound, so that request runs alone and its failure lands in the key'serrors, where_pointsrecords a head failure, instead of aborting the report.generate_recipesrenders every scenario of the named key, soinfx generate --config-key <video key>writes the video server recipes.signoff_facts.matrix_atis unchanged. The CODEOWNER sign-off generates LLM points only, from config data and trusted code, so it never loads a revision'svideo-generation/specs/video_spec.py.generate_config_matrixwithoutscenario_types,full-sweepandtest-configwithout--scenario-typestill leave video out.Results:
Local runs at eb50d8b against origin/main 90bc842, no GPU and no sweep; every row is a local run with no link.
changed --base origin/main --jsonafter editing the H100serving-client-tp2.yamlengine argschanged: [], 184 keysh3-bf16-h100-sglang-video: 3 -> 3 points, 1 added, 1 removed(g2f152db5d9170->c1f805258915), 184 keys, 0 errors, 6.2 s wallchanged --base origin/main --jsonon a clean treechanged: [], 184 keyschanged: [], 184 keys, 0 errors, 6.2 s wallinfx generate --config-key h3-bf16-h100-sglang-videohas no srt-slurm pointsmanifest.jsoninfx.matrix.revision generate <checkout> full-sweep --config-files configs/nvidia-master.yaml configs/amd-master.yaml, with and without--no-evalscmptest-config --config-keys h3-bf16-h100-sglang-video --no-evalswithout--scenario-type[][]signoff_facts.matrix_at("HEAD", workdir)on the real configsTesting:
infx/tests/matrix/test_changed.py::test_keys_with_changed_recipes_or_points_are_listedbuilds a nestedinferencex-e2e/plusvideo-generation/specsrepository with avideokey whose recipe changes between base and head and anunboundkey whose base recipe lacksVIDEO_RUNTIME_REVISION. It expectsvideolisted with one added and one removed point,unboundlisted with one base error, and the text report'swarning:line for that error. With origin/main sources it fails at the base-error assertion (not enough values to unpack); without_base_video_pointsit fails with the base generator'sCalledProcessError; with this branch it passes.infx/tests/srt_slurm/test_generate.py::test_a_video_point_renders_its_bound_server_reciperenders a video key throughinfx generateand checks the bound GPU count, spec path andCONC_LIST. With origin/main sources it fails withinfx: error: video has no srt-slurm points(exit 2); with this branch it passes.infx/tests/workflows/test_signoff_facts.py::test_matrix_at_generates_llm_points_only_and_runs_no_revision_codecommits a repository whosevideo-generation/specs/video_spec.pyraises on import and callsmatrix_at("HEAD", workdir). It expects the LLM key's point, no video points and no errors. It passes on origin/main and on this branch; amatrix_atthat generated video points from the revision snapshot returnederrors == {"video": "video_spec.py from the revision tree ran"}.infx.tests.srt_recipes.write_video_specs, which copies the trustedvideo_spec.pyand writes a one-case schema 1 fixture spec with its own model id and set, so an edit toh3-v2.yamlcannot fail them.infx/tests/suite with-n 4on macOS: 3,243 passed, 2 skipped, 3 failed. The 3 failures are ininfx/tests/matrix/test_revision.py, where the fork server'sos.killpgraisesPermissionError. They fail the same way on a clean origin/main checkout, and this branch does not touchrevision.pyor its tests.ruff checkandruff format --checkpass.Limits:
--scenario-type video-generationrecords that error under each of its video keys; its LLM comparison still completes.matrix_atgenerates LLM scenarios only, as on origin/main.Related: Follows #2916. Pairs with #4885, whose submission guide links the
infx generaterender step this PR fixes for video.PR checklist and change type
Type of Change
Checklist
python -m infx.matrix.changed --base origin/mainlists affected config keys, I have appended one new entry covering them to the physical end ofinferencex-e2e/perf-changelog.yaml; if it lists none, I have added none. I have not edited historical entriesinferencex-e2e/perf-changelog.yamlentry, it carries exactly one primary sweep label (a maintainer applies it on fork PRs):full-sweep-fail-fast(recommended),full-sweep-enabled, ornon-canary-full-sweep-enabled. Optional modifiersall-evals,evals-only, andagentx-fastrequire a primary label; the last two block reuse while applied.OWNER/MEMBER/COLLABORATOR) has commented/use <run_id>(or the legacy/reuse-sweep-run) on this PR. Do this only once there is a final full sweep that is all green with evals passing, since after this comment the primary sweep label will no longer automatically kick off new sweeps. Remove and re-add the primary sweep label to force a new sweep.