TEDPolicy Parameters

Hello,

Could anyone explain to me the role of the parameters of TEDPolicy number_of_transformer_layers Parameter?

In fact, number_of_transformer_layers has the following parameters: text, action_text, label_action_text, dialogue.

What is the meaning of each param? what is the difference between action_text and label_action_text?

Because from this video Rasa Algorithm Whiteboard - TED Policy, I understood that (intent, entity, slot, prev action) arrays are concatenated in a single array. Is this final array corresponding to the dialogue parameter?

I hope you might be able to give me some insights on this.

Thanks!