Modify pattern_clarification

I would like to use the pattern_classification, but modify it to instead use an LLM call to paraphrase the user utterance. I can get this part done using a custom action, but what I would like to do is to actually use the paraphrased sentence to continue the command generation.

To make it more clear, I want to make it something like this:

user: Jimmy said he wanted some cash. send him a hundred rasa: pattern_clarification llm: Transfer hundred dollars to Jimmy rasa: transfer_money_flow

This is an interesting idea, can you elaborate on the underlying problem you’re trying to solve? It sounds like you’re running into unexpected calls to pattern_clarification, is that correct? If so, your first step should be improving your flow descriptions.

I have like super detailed flow descriptions that are a couple of lines themselves. I even added examples. My main issue is that I want this to be working with smaller LLMs. When using GPT-4o I got no issues. But something like phi-3 or gemma2 and it really struggles when the user input is a bit ambiguous. But when I tested with a bunch of examples I could think up, it always ended in pattern_classification. So my idea was to have a detailed set of instructions for the LLM to rephrase the user input, and then use that to identify which flow.

I could see how to implement this when using APIs, but any solution that can be used with inspect and shell would be much appreciated