Skip to content

Record sandbox hook execution metadata #1915

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Harold-lkk wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Record sandbox hook execution metadata #1915

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 19 additions & 0 deletions xtuner/v1/rl/agent_loop/sandbox_agent_loop/sandbox.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -729,6 +729,8 @@ async def run(self, client: Any, item: AgentRolloutItem, record: StageRecord) ->
return result

await self._run_phase("post", self.post, client, item, record, stage_label)
if record.status != StageStatus.FAILED:
record.status = StageStatus.COMPLETED
return result
except Exception as exc:
if record.status != StageStatus.FAILED:
Expand Down Expand Up @@ -824,6 +826,13 @@ async def _run_hook(
hook_name = getattr(hook, "name", hook.__class__.__name__)
task_id = item.id
done = False
hook_record: dict[str, Any] = {
"phase": phase,
"hook": hook_name,
"status": StageStatus.RUNNING.value,
"started_at": started,
}
record.metadata.setdefault("hooks", []).append(hook_record)

async def warn_if_stuck() -> None:
await asyncio.sleep(stuck_warn_sec)
Expand All @@ -835,11 +844,21 @@ async def warn_if_stuck() -> None:
warn_task = asyncio.create_task(warn_if_stuck()) if stuck_warn_sec > 0 else None
try:
await hook(client, item, record)
hook_record["status"] = StageStatus.COMPLETED.value
except Exception as exc:
hook_record["status"] = StageStatus.FAILED.value
hook_record["error"] = {
"type": type(exc).__name__,
"message": str(exc),
}
raise
finally:
done = True
if warn_task is not None:
warn_task.cancel()
elapsed = time.monotonic() - started
hook_record["finished_at"] = time.monotonic()
hook_record["duration_sec"] = elapsed
if elapsed > 1:
get_logger().debug(f"[{task_id}] hook done phase={phase} hook={hook_name} took={elapsed:.1f}s")

Expand Down
Loading