Files
Video-Tree-TRM5/scripts/repair_trees.sh
T
iomgaa 8c383354c5 fix(detector): 移除 ongoing_actions/visible_entities 空字段检测 + 新增 repair sh
抽检确认这两个字段为空是合法内容状态(静物/黑帧/模糊帧),
VLM 重修也修不好,保留会导致断点续跑死循环。
L3 空字段检测缩减为 frame_summary + spatial_layout。
2026-07-09 00:54:27 -04:00

28 lines
918 B
Bash
Executable File
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
#!/usr/bin/env bash
# 树修复管线:检测 + VLM 重生成 + 校验 + Q&A 补全
#
# 用法:
# bash scripts/repair_trees.sh # 默认 16 并发续跑
# bash scripts/repair_trees.sh --dry-run # 仅检测不修复
# bash scripts/repair_trees.sh --reaggregate-all # 强制全量重聚合
# CONCURRENCY=8 bash scripts/repair_trees.sh # 自定义并发数
#
# 日志输出:
# stderr → 终端实时显示
# logs/repair_trees.log → 完整日志(自动 rotation 50MB
# logs/repair_telemetry.db → LLM/VLM 调用遥测
# logs/repair_progress.json → 断点续跑进度
set -euo pipefail
CONCURRENCY="${CONCURRENCY:-16}"
conda activate Video-Tree-TRM
python tools/repair_trees.py \
--videos-dir store/videos \
--srt-dir data/Video-MME/subtitle \
--questions-dir store/questions/benchmarks/Video-MME \
--concurrency "$CONCURRENCY" \
"$@"