Repository navigation
Validate current XPalm benchmarks and release unused samples - #222
Merged
Merged
Conversation
VEZY
changed the title
Contributor
Benchmark Results (Julia v1.12.1)Time benchmarks
Memory benchmarks
|
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The current XPalm model was compared with a historical 0.6.1 final LAI. Use the released
v0.7.0regression reference after validating its source commit, default-parameter scenario, weather hash, dates, and complete 4,160-day horizon. Historical release comparisons, numerical fixture data, and tolerances remain unchanged.Repeated full-output measurements previously kept large histories alive between samples. Extract the first run's final state and runtime counters outside the timed operation, then release its history before the next sample. Non-inlined sample helpers also release later results and factory-owned model state. Default callers still receive the first result unchanged; lifetime tests cover both paths, sample counts, and statistics.
The full workflow retains the complete output-retention matrix with three samples per stage. The optional warmed no-output diagnostic remains separate. Progress/resource logging and a bounded test step preserve checkpoint and artifact-upload time.
XPalm compatibility PR #134 and reference-promotion PR #137 are merged. Integration and the full benchmark use XPalm
main. The promotion moves the existing numerical CSV blobs unchanged and preserves their original generation provenance. PlantSimEngine PR #223 is included in this head.Validation at
deca45a12a54867fbe856da4f078a58019bc07e6:xpalm_ref=main,plantbiophysics_ref=master,plantgeom_ref=main, andxpalm_profile=full. PlantBiophysics API smoke: 14/14; performance: 5/5; XPalm oracle: 12/12; measurement lifetime: 30/30; complete profile: 4/4.689d9117b6ccb51bc64e4603b280b68c0ad9ed5f(main), PlantBiophysics9ee42793c54118111a0b72dbfa9bc04f4ccd6422(master), and PlantGeom2933d88b3acb2261f9591d4257b0024434095ea0(main). The manifest, fixture hash, source revisions, measurements, and resource log are retained in the workflow artifact.znver5, and the recorded warmup policy. Maximum sampled process RSS was 11.881 GiB. Timings describe this resolved environment and full-profile measurement method.All current checks pass: engine CI across Julia 1.10/current on Linux, macOS, and Windows; downstream integration for XPalm, PlantBiophysics, and PlantGeom; regular benchmarks; documentation/deploy; browser E2E; and the manually dispatched full matrix above. There are 14 successful check runs; the ordinary PR workflow skips its optional manual full-performance job, which is covered by the successful explicit run. Independent artifact verification confirms every stage/sample count, final state, revision, and manifest hash.
git diff --checkpasses.