You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It would be great if there could be some way that Devon outputs a history so that if the workflow is stopped it can be resumed again upon reloading. I Have tried a few tricks to keep my workflow going but I still keep hitting rate limits. My attempt was to create an instructions.txt with a very detailed explanation of what I wanted. In my instructions I told Devon to keep a running log in a separate txt file so that the workflow could be easily restarted in case of a crash. Within my instructions.txt I tell Devon to first read the previous log and then continue working. It kind of works, but not well enough to get me through the entire project.
The text was updated successfully, but these errors were encountered:
@basher83 Yup, better persistence / restoring from a history is definitely on the list, in order to make it a lot more usable. In the meantime, a quick trick some have done just to avoid the per-minute rate limit is add a sleep duration in the code.
That's a actually really cool way to persist workflows!Adding persistence is definitely on the roadmap.
Rate limit is strange as I added exponential backoff. I'll increase backoff period so those can be avoided!
@ObjectJosh Thanks for the tip! I suppose a sleep duration would have been a simpler approach to my complex solution. I appreciate the feedback and look forward to how this project improves and evolves.
It would be great if there could be some way that Devon outputs a history so that if the workflow is stopped it can be resumed again upon reloading. I Have tried a few tricks to keep my workflow going but I still keep hitting rate limits. My attempt was to create an instructions.txt with a very detailed explanation of what I wanted. In my instructions I told Devon to keep a running log in a separate txt file so that the workflow could be easily restarted in case of a crash. Within my instructions.txt I tell Devon to first read the previous log and then continue working. It kind of works, but not well enough to get me through the entire project.
The text was updated successfully, but these errors were encountered: