From 933de6251e25f5b99bb4afc3c1808fbeed9d84ce Mon Sep 17 00:00:00 2001
From: iomgaa <iomgaaycz@gmail.com>
Date: Thu, 16 Jul 2026 07:46:31 -0400
Subject: [PATCH] chore: add FORCE flag to split script; align train config
 pool sizes to 0.4 split

---
 config/train_videomme.yaml   | 5 +++--
 scripts/build_video_split.sh | 9 ++++++++-
 2 files changed, 11 insertions(+), 3 deletions(-)

diff --git a/config/train_videomme.yaml b/config/train_videomme.yaml
index 83380a4..8be6868 100644
--- a/config/train_videomme.yaml
+++ b/config/train_videomme.yaml
@@ -40,10 +40,11 @@ harness:
   skill_update_mode: patch
   appendix_consolidate_threshold: 6
   # 池配置 —— global 冻结切分（diag/val/test 尺寸由 pools.json 冻结，以下采样旋钮加载时忽略）
+  # 实际冻结切分（0.4+tier，fingerprint d456ef4f）: diag=180 val=120 test=600
   pool_split_mode: global
-  diag_size: 210
+  diag_size: 180
   diag_correct_ratio: 0.5
-  val_size: 90
+  val_size: 120
   val_correct_ratio: 0.5
   test_size: 600
   test_questions: "benchmarks/Video-MME"
diff --git a/scripts/build_video_split.sh b/scripts/build_video_split.sh
index ab40559..90c1b77 100755
--- a/scripts/build_video_split.sh
+++ b/scripts/build_video_split.sh
@@ -50,6 +50,13 @@ if [ "${MODE:-}" = "mock" ]; then
   DRY_RUN_FLAG="--dry-run"
 fi
 
+# FORCE=1 → --force：已存在冻结产物时覆盖（旧产物自动备份为 .bak.*，WP2 覆盖保护）。
+# 重跑切分（改 val_ratio / tier 感知 / 换指纹）时需要；缺省拒绝覆盖以防误毁冻结锚点。
+FORCE_FLAG=""
+if [ "${FORCE:-}" = "1" ]; then
+  FORCE_FLAG="--force"
+fi
+
 # 诊断并发上限（asyncio.Semaphore）：默认 12，可用 CONCURRENCY 环境变量覆盖。
 # 注：LLM 熔断阈值/冷却是工程配置，走 .env（LLM_CIRCUIT_BREAKER_THRESHOLD / _COOLDOWN），
 #     当前 threshold=32（连续 32 次失败触发熔断），不在本脚本设置。
@@ -58,4 +65,4 @@ CONCURRENCY="${CONCURRENCY:-12}"
 # conda run --no-capture-output：不需 conda init（非交互脚本可跑），且实时直出不缓冲。
 conda run --no-capture-output -n Video-Tree-TRM python -u -m app.harness.video_split_cli \
   --config config/video_split.yaml \
-  --concurrency "${CONCURRENCY}" ${DRY_RUN_FLAG}
+  --concurrency "${CONCURRENCY}" ${DRY_RUN_FLAG} ${FORCE_FLAG}