Does anyone know how to resolve this issue or if it’s even possible to work around it? I was working with a notebook where I used llm=OpenAI(model=gpt-4), then I tried with llama3-70b-8192 from Groq, and encountered this problem.