File size: 3,007 Bytes
f15a766 | 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 | from __future__ import annotations
import argparse
import json
from pathlib import Path
import numpy as np
def summarize_variant(root: Path, tag: str) -> dict:
ckpt_dir = root / tag / "checkpoint-100000"
aggregate_files = sorted(ckpt_dir.glob("val_psnr_*.json"))
if aggregate_files:
data = json.loads(aggregate_files[-1].read_text(encoding="utf-8"))
return {
"tag": tag,
"samples": data.get("num_samples", 0),
"psnr_mean": data.get("psnr_mean"),
"psnr_std": data.get("psnr_std"),
"psnr_min": data.get("psnr_min"),
"psnr_max": data.get("psnr_max"),
"complete": True,
}
metrics_paths = sorted(ckpt_dir.glob("val*/metrics.json"))
psnr_values = []
for path in metrics_paths:
data = json.loads(path.read_text(encoding="utf-8"))
value = data.get("psnr_mean")
if value is not None and np.isfinite(value):
psnr_values.append(float(value))
if not psnr_values:
return {
"tag": tag,
"samples": 0,
"psnr_mean": None,
"psnr_std": None,
"psnr_min": None,
"psnr_max": None,
"complete": False,
}
values = np.array(psnr_values, dtype=np.float64)
return {
"tag": tag,
"samples": int(values.size),
"psnr_mean": float(values.mean()),
"psnr_std": float(values.std()),
"psnr_min": float(values.min()),
"psnr_max": float(values.max()),
"complete": False,
}
def main() -> None:
parser = argparse.ArgumentParser(description="Summarize query val PSNR eval outputs.")
parser.add_argument("--root", type=Path, default=Path("/workspace/Ctrl-World-Graph/eval_query_val"))
parser.add_argument(
"--tags",
nargs="*",
default=[
"gps_query",
"gine_query",
"gatv2_query",
"transformer_query",
"edge_transformer_query",
"hybrid_gine_transformer_query",
],
)
parser.add_argument("--out", type=Path, default=None)
args = parser.parse_args()
rows = [summarize_variant(args.root, tag) for tag in args.tags]
rows.sort(key=lambda row: float("-inf") if row["psnr_mean"] is None else row["psnr_mean"], reverse=True)
header = "tag\tsamples\tcomplete\tpsnr_mean\tpsnr_std\tpsnr_min\tpsnr_max"
lines = [header]
for row in rows:
fmt = lambda value: "NA" if value is None else f"{value:.4f}"
lines.append(
f"{row['tag']}\t{row['samples']}\t{row['complete']}\t"
f"{fmt(row['psnr_mean'])}\t{fmt(row['psnr_std'])}\t"
f"{fmt(row['psnr_min'])}\t{fmt(row['psnr_max'])}"
)
text = "\n".join(lines)
print(text)
if args.out is not None:
args.out.parent.mkdir(parents=True, exist_ok=True)
args.out.write_text(text + "\n", encoding="utf-8")
if __name__ == "__main__":
main()
|