mirror of
https://github.com/manifoldbt/manifoldbt.git
synced 2026-11-02 07:56:11 +00:00
Jimmy7892
7e16dcdbb3
GitHub's hosted runners have no GPU, so the sweeps job only ever measures the CPU path. A new workflow, bench-gpu.yml, runs the same three grids, on the same data and the same strategy imported from vs_vectorbt, on a Modal L4, with the CPU of that machine beside it on 4 reserved cores. The harness fails a point when the GPU sweep falls back to the CPU (the engine only warns), when any cell differs from the CPU grid beyond the gate the CPU job uses, when a repetition does not reproduce the first call, or when the licence tier changes. It calibrates the card and runs each point once untimed before timing, and reports that first call apart. The result is published to benchmarks/gpu/results/latest.json, separate from the CPU file, only for a green run of the latest version on the L4. Without a Modal token the job warns and measures nothing. Checked on a workstation (RTX 3090, PyPI wheel 0.28.1): every cell bit-identical to the CPU on all three points; with the card hidden, a run stops at calibration, and a sweep forced onto the CPU raises the warning the harness catches (the verdict it then gives is covered by the tests); 19 offline tests pass and actionlint is clean. Not yet run on Modal: that needs the token.
67 lines
2.5 KiB
Python
67 lines
2.5 KiB
Python
"""Wrap a GPU run's result with its provenance, as the file the website reads.
|
|
|
|
python publish.py <artifact-dir> --out results/latest.json
|
|
|
|
Same contract as ../vs_vectorbt/publish.py: the numbers are committed to the
|
|
repository rather than left in an artifact, which needs a token to download and
|
|
expires after ninety days.
|
|
|
|
The GPU result is a file of its own rather than a third key in the CPU file. It
|
|
comes from another machine, in another job, and a card that could not be had
|
|
must not hold back the CPU numbers, nor the reverse.
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import argparse
|
|
import json
|
|
import os
|
|
import sys
|
|
from datetime import datetime, timezone
|
|
from pathlib import Path
|
|
|
|
SOURCE = "results-gpu.json"
|
|
|
|
|
|
def main() -> int:
|
|
parser = argparse.ArgumentParser(description=__doc__)
|
|
parser.add_argument("artifacts", help="directory the artifact was downloaded into")
|
|
parser.add_argument("--out", required=True, help="file to write")
|
|
args = parser.parse_args()
|
|
|
|
found = next(iter(sorted(Path(args.artifacts).rglob(SOURCE))), None)
|
|
if found is None:
|
|
print("no {} under {}".format(SOURCE, args.artifacts), file=sys.stderr)
|
|
return 1
|
|
payload = json.loads(found.read_text(encoding="utf-8"))
|
|
# The publish job only runs after a green measuring job, and the harness
|
|
# exits non-zero on any failed point. Checked again here because this file
|
|
# is what the website renders.
|
|
if payload.get("status") != "exact":
|
|
print("refusing to publish a result whose status is {!r}".format(
|
|
payload.get("status")), file=sys.stderr)
|
|
return 1
|
|
|
|
server = os.environ.get("GITHUB_SERVER_URL", "https://github.com")
|
|
repo = os.environ.get("GITHUB_REPOSITORY", "manifoldbt/manifoldbt")
|
|
run_id = os.environ.get("GITHUB_RUN_ID", "")
|
|
published = {
|
|
"repo": repo,
|
|
"run_id": run_id,
|
|
"run_url": "{}/{}/actions/runs/{}".format(server, repo, run_id) if run_id else "",
|
|
"commit": os.environ.get("GITHUB_SHA", ""),
|
|
"run_started_at": os.environ.get("GITHUB_RUN_STARTED_AT", ""),
|
|
"conclusion": "success",
|
|
"synced_at": datetime.now(timezone.utc).isoformat(),
|
|
"gpu": payload,
|
|
}
|
|
|
|
out = Path(args.out)
|
|
out.parent.mkdir(parents=True, exist_ok=True)
|
|
out.write_text(json.dumps(published, indent=2) + "\n", encoding="utf-8")
|
|
print("wrote {} ({:.0f} KB)".format(out, out.stat().st_size / 1024))
|
|
return 0
|
|
|
|
|
|
if __name__ == "__main__":
|
|
raise SystemExit(main())
|