Self-Forcing-part-2 / scripts /wait_and_sync_predictor_v4_artifacts.sh
Cccccz's picture
Add files using upload-large-folder tool
2847d0b verified
Raw
History Blame Contribute Delete
839 Bytes
#!/usr/bin/env bash
set -euo pipefail
ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
RUN_DIR="${RUN_DIR:-$ROOT_DIR/checkpoints/predictor_v4_1000_seed0}"
POLL_SECONDS="${POLL_SECONDS:-60}"
UPSTREAM_PID="${UPSTREAM_PID:-}"
if [[ -z "$UPSTREAM_PID" ]]; then
if [[ ! -f "$RUN_DIR/end_to_end.pid" ]]; then
echo "错误:缺少 $RUN_DIR/end_to_end.pid。" >&2
exit 1
fi
UPSTREAM_PID="$(cat "$RUN_DIR/end_to_end.pid")"
fi
if [[ ! "$UPSTREAM_PID" =~ ^[1-9][0-9]*$ ]]; then
echo "错误:UPSTREAM_PID 不是有效 PID:$UPSTREAM_PID" >&2
exit 1
fi
echo "等待 end-to-end supervisor PID=$UPSTREAM_PID 完成。"
while kill -0 "$UPSTREAM_PID" 2>/dev/null; do
sleep "$POLL_SECONDS"
done
echo "Supervisor 已退出,检查完整产物后开始同步。"
bash "$ROOT_DIR/scripts/sync_predictor_v4_artifacts.sh"