human_input for tasks not working as expected
I am trying to use crew ai for a travel booking use case I have an agent with a task assigned to it. The description of the task is to gather information from the human and I want it to be as interactive as possible. I set the human_input=True for that task, but the agent assumes some random information and only then asks for human input at the end of the task.
My question is: How to het a human input in the middle of the task? Do I refine my prompt? Should I give the task a better description? Should I give the tasks a better expected_output? Can I use any other task attribute to make the human_input more interactive
@rohanprasadsap Did you find anything? I have having the same issue and encountering an error