Closed evangriffiths closed 1 month ago
Here's the commit where the training feature is added https://github.com/crewAIInc/crewAI/commit/175d5b3dd6a37b54708a89bcca64df2122f2b7c9.
My understanding of how it works:
crew.train
turns this into an iterative process, where it runs the agent normally, with the 'human feedback' feature enabled, but...train
process, it loads the contents of the second .pkl file into the agent's promptAll-in-all, this is quite similar to what we were trying to do with RememberPastActions
, but with important differences:
RememberPastActions
, but here the human gives feedback every iteration, and this is appended to the prompt in a fixed way.I haven't tried crewai.train, so don't know if it works that well, but I think we could take inspiration from this for the general agent for sure.
Overlaps with https://github.com/gnosis/prediction-market-agent/issues/207. Think about these issues together
I've written up an idea for how to apply the learnings from this feature in a new issue here https://github.com/gnosis/prediction-market-agent/issues/370, so closing this issue.
https://docs.crewai.com/core-concepts/Training-Crew
Have only had a glance so far. Looks like it does some form of RL to tune the prompts! This is exactly the kind of feature we want for our PM agents to learn over time!