(feat) Add trajectory replay for headless mode #6215
Conversation
|
Demo: Step 1. Run OpenHands and generate a trajectory workspace_base="./workspace"
save_trajectory_path="./traj/demo.json"With command: Step 2. Run OpenHands with trajectory replay workspace_base="./workspace2"
replay_trajectory_path="./traj/demo.json"With command: Step 3. Check the result 
|
| @@ -108,6 +109,7 @@ def __init__( | |||
| is_delegate: Whether this controller is a delegate. | |||
| headless_mode: Whether the agent is run in headless mode. | |||
| status_callback: Optional callback function to handle status updates. | |||
| replay_logs: A list of logs to replay. | |||
There was a problem hiding this comment.
How about we call it replay_events, since it is really a list of events
| self.replay_logs = replay_logs | ||
| self.replay_index = 0 | ||
| if self.replay_logs: | ||
| logger.info(f'Replay logs loaded, events length = {len(self.replay_logs)}') |
There was a problem hiding this comment.
Can we consolidate these replay related thing to a separate class? eg something like StuckDetector to help reduce the complexity of AgentController?
There was a problem hiding this comment.
I agree, and it's a very self-contained feature
There was a problem hiding this comment.
Although, it does execute the actions, right?
End-user friendly description of the problem this fixes or functionality that this introduces
Add trajectory replay feature for headless mode.
Give a summary of what the PR does, explaining any non-trivial design decisions
To test it out, add
replay_trajectory_pathtoconfig.toml, and thenTODOs in this PR:
trajectories_pathconfig tosave_trajectory_path#6216 is mergedadd tests(non-trivial, will skip in this PR)TODOs in next PR:
Link of any specific issues this addresses
#6049
To run this PR locally, use the following command: