How can I convert output.jsonl into a format similar to SWE-Gym/OpenHands-SFT-Trajectories?

[lycfight]

What problem or use case are you trying to solve?

Describe the UX of the solution you'd like

Do you have thoughts on the technical implementation?

Describe alternatives you've considered

Additional context

@xingyaoww

I want to train a model using long dialogue trajectory data generated by OpenHands to improve performance. However, I am confused by the output.jsonl data generated by run_infer. Some history entries have an odd number of turns, some have an even number, and some rounds are empty. This differs significantly from the common multi-turn dialogue SFT data format.

Which parts of the data can be used to form a proper multi-turn dialogue training format? Can you explain the meaning of the fields in this file in detail? Also, how can I convert output.jsonl into a format similar to SWE-Gym/OpenHands-SFT-Trajectories?