simonkoson
96 lines
3.1 KiB
Python
96 lines
3.1 KiB
Python
"""
|
|
summarize.py - 汇总打标结果命中情况
|
|
用法: python summarize.py --model m3
|
|
"""
|
|
|
|
import sys
|
|
sys.stdout.reconfigure(encoding='utf-8')
|
|
sys.stderr.reconfigure(encoding='utf-8')
|
|
|
|
import json
|
|
import glob
|
|
import argparse
|
|
from pathlib import Path
|
|
|
|
BASE_DIR = Path(__file__).parent.parent
|
|
EXPERIMENTS_DIR = BASE_DIR / "experiments"
|
|
|
|
|
|
def load_json(path):
|
|
try:
|
|
return json.loads(path.read_text(encoding="utf-8"))
|
|
except Exception:
|
|
return None
|
|
|
|
|
|
def latest_per_ep(files):
|
|
"""同 ep 有多个文件时取时间戳最新的。"""
|
|
latest = {}
|
|
for f in files:
|
|
name = f.name
|
|
# 文件名格式: 20260611_154037_m3_ep04.json
|
|
parts = name.replace(".json", "").split("_")
|
|
if len(parts) >= 4:
|
|
ts = parts[0] + parts[1] # yyyymmddHHMMSS
|
|
ep = int(parts[-1].replace("ep", ""))
|
|
key = ep
|
|
if key not in latest or ts > latest[key][1]:
|
|
latest[key] = (f, ts)
|
|
return {ep: info[0] for ep, info in latest.items()}
|
|
|
|
|
|
def run(model):
|
|
pattern = str(EXPERIMENTS_DIR / f"*_{model}_*.json")
|
|
files = sorted(Path(p) for p in glob.glob(pattern))
|
|
if not files:
|
|
print(f"未找到 {pattern}")
|
|
return
|
|
|
|
ep_files = latest_per_ep(files)
|
|
rows = []
|
|
parse_fail = 0
|
|
|
|
for ep in sorted(ep_files.keys()):
|
|
data = load_json(ep_files[ep])
|
|
if data is None:
|
|
parse_fail += 1
|
|
rows.append({"ep": ep, "title": "?", "gt": "?", "pred": "解析失败", "hit": False, "conf": "?"})
|
|
continue
|
|
|
|
gt = data.get("ground_truth", {})
|
|
result = data.get("result")
|
|
|
|
title = gt.get("title", "?")
|
|
gt_val = gt.get("narrative_structure", "?")
|
|
pred_val = result.get("narrative_structure") if result else None
|
|
conf = result.get("confidence", "?") if result else "?"
|
|
hit = pred_val == gt_val if pred_val is not None else False
|
|
|
|
rows.append({"ep": ep, "title": title, "gt": gt_val, "pred": pred_val or "解析失败", "hit": hit, "conf": conf})
|
|
|
|
# 打印每行
|
|
for r in rows:
|
|
mark = "✓" if r["hit"] else "✗"
|
|
conf_str = f'置信度:{r["conf"]}' if r["conf"] != "?" else ""
|
|
print(f' ep{r["ep"]:02d} {r["title"]:<10} | 标准:{r["gt"]:<8} | {model}:{r["pred"]:<8} | {mark} {conf_str}')
|
|
|
|
# 汇总
|
|
total = len(rows)
|
|
hits = sum(1 for r in rows if r["hit"])
|
|
hi_conf = [r for r in rows if r["conf"] == "高"]
|
|
mid_low = [r for r in rows if r["conf"] in ("中", "低")]
|
|
hi_hit = sum(1 for r in hi_conf if r["hit"])
|
|
ml_hit = sum(1 for r in mid_low if r["hit"])
|
|
|
|
print(f"\n ===== {model} 命中情况 =====")
|
|
print(f" narrative_structure 命中: {hits}/{total} = {hits*100//total}%")
|
|
print(f" 自评\"高\"置信的命中率: {hi_hit}/{len(hi_conf)}")
|
|
print(f" 自评\"中/低\"置信的命中率: {ml_hit}/{len(mid_low)}")
|
|
print(f" 解析失败: {parse_fail} 期")
|
|
|
|
|
|
if __name__ == "__main__":
|
|
parser = argparse.ArgumentParser()
|
|
parser.add_argument("--model", required=True, help="模型键名，如 m3 / deepseek-v4")
|
|
args = parser.parse_args()
|
|
run(args.model) |