File size: 5,674 Bytes
9422383 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 | #!/usr/bin/env python3
"""Record review evidence around an unchanged Ultralytics benchmark command.
Ultralytics remains responsible for loading the dataset, running validation,
and reporting accuracy and stage timings. This command retains that output with
the measured runtime and model identities, source revision, and Raspberry Pi state needed to
interpret the result, then exits with the benchmark's status.
"""
from __future__ import annotations
import argparse
import hashlib
import json
import resource
import shutil
import subprocess
import sys
import tempfile
from collections.abc import Mapping, Sequence
from datetime import UTC, datetime
from pathlib import Path
from .host_metrics import capture_host_snapshot
__all__ = ["record_benchmark"]
def _sha256(path: Path) -> str:
digest = hashlib.sha256()
with path.open("rb") as stream:
for block in iter(lambda: stream.read(1024 * 1024), b""):
digest.update(block)
return digest.hexdigest()
def _run(command: Sequence[str], log_path: Path) -> tuple[int, dict[str, object]]:
"""Run once, preserving output and descriptive state from both boundaries."""
started_at = datetime.now(UTC)
before = capture_host_snapshot()
with log_path.open("x", encoding="utf-8") as log:
process = subprocess.Popen(
command,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
text=True,
encoding="utf-8",
errors="replace",
bufsize=1,
)
if process.stdout is None:
raise RuntimeError("benchmark output stream is unavailable")
for line in process.stdout:
sys.stdout.write(line)
sys.stdout.flush()
log.write(line)
log.flush()
return_code = process.wait()
after = capture_host_snapshot()
finished_at = datetime.now(UTC)
child_usage = resource.getrusage(resource.RUSAGE_CHILDREN)
report: dict[str, object] = {
"version": 1,
"kind": "yolo26n-ultralytics-benchmark-result",
"command": list(command),
"started_at": started_at.isoformat(),
"finished_at": finished_at.isoformat(),
"exit_code": return_code,
# Linux reports ru_maxrss in KiB. The fresh recorder process has only
# one child command, so this is the benchmark process tree's peak.
"child_peak_rss_bytes": child_usage.ru_maxrss * 1024,
"host": {"before": before.to_json(), "after": after.to_json()},
}
return return_code, report
def _write_json(path: Path, value: Mapping[str, object]) -> None:
with path.open("x", encoding="utf-8") as stream:
json.dump(value, stream, indent=2, sort_keys=True)
stream.write("\n")
def record_benchmark(
output: Path,
runtime_library: Path,
model_manifest: Path,
revision: str,
command: Sequence[str],
) -> int:
"""Publish one complete measurement directory without replacing prior evidence."""
if output.exists():
raise FileExistsError(f"output already exists: {output}")
if not runtime_library.is_file():
raise FileNotFoundError(f"runtime library does not exist: {runtime_library}")
if not model_manifest.is_file():
raise FileNotFoundError(f"model manifest does not exist: {model_manifest}")
manifest_document = json.loads(model_manifest.read_text(encoding="utf-8"))
if not isinstance(manifest_document, Mapping):
raise ValueError("model manifest must be a JSON object")
output.parent.mkdir(parents=True, exist_ok=True)
staging = Path(tempfile.mkdtemp(prefix=f".{output.name}.", dir=output.parent))
try:
return_code, result = _run(command, staging / "benchmark.log")
result["source_revision"] = revision
result["runtime_library"] = {
"path": str(runtime_library),
"bytes": runtime_library.stat().st_size,
"sha256": _sha256(runtime_library),
}
result["model_manifest"] = {
"path": str(model_manifest),
"sha256": _sha256(model_manifest),
"document": manifest_document,
}
_write_json(staging / "results.json", result)
staging.rename(output)
finally:
shutil.rmtree(staging, ignore_errors=True)
return return_code
def _main() -> None:
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("--output-dir", type=Path, required=True)
parser.add_argument("--runtime-library", type=Path, required=True)
parser.add_argument("--model-manifest", type=Path, required=True)
parser.add_argument("--revision", required=True)
parser.add_argument("command", nargs=argparse.REMAINDER)
args = parser.parse_args()
command = args.command[1:] if args.command[:1] == ["--"] else args.command
if not command:
parser.error("a benchmark command is required after --")
revision = args.revision.strip()
if not revision or "\n" in revision:
parser.error("--revision must be one non-empty line")
output = args.output_dir.expanduser().resolve()
if output.exists():
parser.error(f"output already exists: {output}")
return_code = record_benchmark(
output,
args.runtime_library.expanduser().resolve(),
args.model_manifest.expanduser().resolve(),
revision,
command,
)
print(f"benchmark evidence: {output}")
if return_code:
raise SystemExit(return_code)
if __name__ == "__main__":
try:
_main()
except (OSError, RuntimeError, ValueError) as error:
raise SystemExit(f"error: {error}") from None
|