Any idea when the integration with

we mostly use langchain's llm abstractions, might be a good question for their discord

If there's interest though we could always subclass our own llmpredictor