I’m using rasa open source , usually my chatbot takes 2 seconds to answer , but since i’m using action server , it takes about 6 seconds to answer , even answers that doesn’t envolve using action server takes also long time , i want to know if there is a solution or best practises to reduce this time .
The action server is not called unless a custom action needs to be called so there must be something else that is changed. I would review the debug lot to see where the time is being spent. You could run it thru my rasalog script to help evaluate the log.
Can you please provide me the steps with more expalanation please ?
Run Rasa with the --debug option.
After debugging :
2024-03-19 09:55:14 DEBUG rasa.core.processor - Logged UserUtterance - tracker now has 4 events.
2024-03-19 09:55:14 DEBUG rasa.core.actions.action - Calling action endpoint to run action 'action_problem'.
2024-03-19 09:55:16 DEBUG rasa.core.actions.action - Calling action endpoint to run action 'action_city'.
2024-03-19 09:55:18 DEBUG rasa.core.actions.action - Validating extracted slots:
2024-03-19 09:55:18 DEBUG rasa.core.processor - [debug ] processor.extract.slots action_extract_slot=action_extract_slots len_extraction_events=0 rasa_events=[]
, i was just saying hi , and the chatbot try the call all the actions that doesn't need , i want to know if there is a solution for this problem
The more actions implemented in an action file, the more time-consuming it becomes to process all intent answers.
Can you share your repo in github include the action server source code and rule/story/form related to the above log?