While performing the rasa train, getting the below error
lib\site-packages\rasa\core\policies\ted_policy.py", line 876, in _check_data raise ValueError( ValueError: No user features specified. Cannot train ‘TED’ model.
It may be because user intents are absent from your stories. Try beginning the stories with something like intent: greet and see if it works.
got it, Thank You… I want to start with actions only (only by clicking buttons) instead of typing. When i start the story with action always in rasa shell asks for your input -> how to do this?
I think Rasa always waits for user input to start the conversation. You could begin your stories with
- intent: session_start
- action: action_session_start
if you wanted Rasa to start the conversation. I’m not totally sure about that though.
Thanks Alex, my requirement is if i open the chatbot by default Rasa should mention Hi message.
I am actually not sure how to get Rasa to initiate the conversation. Maybe someone else can give an answer or maybe you can create a new post for that.
Ok. Thank You.
Hey, @santhoshb Could you please share the solution if you were able to get it
It seems like the error you’re encountering during the Rasa training process is due to missing user features. The TED (Transformer Embedding Dialogue) policy requires user features to be specified for training, but it appears that they are not provided in your data.
To resolve this issue, you’ll need to make sure that your training data includes the necessary user features required by the TED policy. These features typically include information about the users, such as their intents, entities, and other relevant contextual information.
Once you’ve ensured that your training data contains the required user features, you should be able to train the TED model without encountering this error. If you need further assistance, feel free to provide more details about your training data and configuration, and I can help you troubleshoot further.