#!/bin/bash
# 核对各条 OPD 线实际生效的 train_batch / ppo_mini_batch / rollout.n
D=/home/haojinw2
echo "=== 从启动脚本读"
for f in opd_resp7168_flat.sh chain113.sh chain_fp32.sh chain_justrl_mb16.sh \
         rose-bundle-0808/run_opd_baseline.sh rose-bundle-0808/run_opd_official.sh \
         OPD-upstream/on_policy_distillation.sh; do
  p=$D/$f
  [ -f "$p" ] || continue
  echo "  --- $f"
  grep -aoE "(data\.train_batch_size|ppo_mini_batch_size|rollout\.n|TRAIN_BATCH_SIZE|MINI_BATCH_SIZE|PPO_MINI_BATCH_SIZE|N_RESPONSES)=[^ \\\\]*" "$p" 2>/dev/null | sed 's/^/      /'
done
echo
echo "=== 从 wandb 离线配置读实际生效值(最可靠)"
for d in $D/rose-bundle-0808/verl/wandb/offline-run-*; do
  c="$d/files/config.yaml"
  [ -f "$c" ] || continue
  name=$(grep -aoE "experiment_name:.*" "$c" | head -1 | sed 's/experiment_name: *//' | tr -d '"')
  [ -z "$name" ] && continue
  tb=$(grep -aA2 "^ *train_batch_size:" "$c" | grep -aoE "value: [0-9]+" | head -1 | grep -oE "[0-9]+")
  mb=$(grep -aA2 "ppo_mini_batch_size:" "$c" | grep -aoE "value: [0-9]+" | head -1 | grep -oE "[0-9]+")
  nn=$(grep -aA2 "^ *n:" "$c" | grep -aoE "value: [0-9]+" | head -1 | grep -oE "[0-9]+")
  printf "  %-46s train=%-5s mini=%-5s n=%-3s\n" "${name:0:46}" "${tb:-?}" "${mb:-?}" "${nn:-?}"
done | sort -u