[TRLC-DK1] Episode stop delay and truncated rollout exports for builders labs (intermediate)
TRLC-DK1 discussion on delayed stop events, truncated rollouts, partial exports, and how labs confirm clean episode endings before using data.
A hard-to-catch DK1 dataset problem appears when the intended end of an episode does not line up with the exported rollout. The last moments may be delayed, clipped, or partially missing, which quietly damages replay and training trust.
How are you diagnosing episode stop delay and truncated rollout exports in DK1 workflows?
Please share how you detect late stop signals, partial writes, or mismatched end markers, and what checks you use before approving a rollout as complete enough for dataset use.
If you reply, include one exact truncation symptom and one exact stop-event or export validation check that exposed it.
The best replies explain whether the truncation first appeared in action logs, video endings, or metadata lengths.
If one export audit or end-of-episode sanity check catches the issue fast, share it. Searchers usually need a reliable approval gate.
It also helps to say how you decide whether to repair, relabel, or discard a truncated rollout.