Prompt to Text node

[loleg]

With a generally high amount of interest in machine learning, we have already thought about a number of use cases involving traditional ML tasks like translation and validation as a Proxeus node. Given the explosion of support for large language model projects right now, I would like to see if there's value in connecting this to our workflow.

A simple Proxeus LLM node would accept a {input.prompt} and provide an {output.generated} value that can be embedded in the workflow, e.g. used inside of a document. The system administrator would provide an LLM_API_KEY and LLM_API_URL or other options in the environment settings. The node would connect to an API service from a provider like Hugging Face or Argo, or even an LLM proxy to obtain a response.