fix(eval): 单条样本失败不再拖垮整轮评估

- cli：逐样本 transcribe 包 try/except，失败记为该样本 error 并继续；
  失败样本 counts 置空，不计入准确率，仅在 errors 列计数
- custom 数据集：manifest 引用的音频不存在时 warn 跳过，不再让 soundfile
  抛错崩掉整轮（修复默认 config 的 custom 示例集指向不存在音频导致的崩溃）

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

eval/asr_eval/cli.py

@@ -17,8 +17,8 @@ from rich.console import Console
 from rich.progress import BarColumn, Progress, TextColumn, TimeElapsedColumn
 
 from .datasets import build_dataset
-from .engines import build_engine
+from .engines import Transcript, build_engine
-from .metrics.err import score_sample
+from .metrics.err import empty_counts, score_sample
 from .metrics.resource import ResourceProbe
 from .report import aggregate, render_markdown, write_reports
 
@@ -117,8 +117,12 @@ def cmd_run(args: argparse.Namespace) -> int:
         ) as prog:
             task = prog.add_task(f"{name}", total=len(samples))
             for s in samples:
-                tr = _transcribe_one(engine, s.audio_path)
+                try:
-                sc = score_sample(s.ref_text, tr.text)
+                    tr = _transcribe_one(engine, s.audio_path)
+                except Exception as ex:  # 单条样本失败不拖垮整轮评估
+                    tr = Transcript(text="", audio_sec=0.0, proc_sec=0.0, error=str(ex))
+                # 失败样本不计入准确率（counts 置空），仅在 errors 列计数
+                sc = empty_counts() if tr.error else score_sample(s.ref_text, tr.text)
                 rows.append({
                     "engine": name,
                     "dataset": s.dataset,

eval/asr_eval/datasets/custom.py

@@ -11,6 +11,7 @@ from __future__ import annotations
 import json
 import os
 import re
+import sys
 from collections.abc import Iterator
 
 from ..manifest import Sample
@@ -50,6 +51,9 @@ class CustomDataset:
                 audio = row["audio"]
                 if not os.path.isabs(audio):
                     audio = os.path.join(self._base, audio)
+                if not os.path.exists(audio):
+                    print(f"[custom] 跳过缺失音频：{audio}", file=sys.stderr)
+                    continue
                 lang = row.get("lang") or _guess_lang(text)
                 n += 1
                 yield Sample(