Prompt, Generate, then Cache: Cascade of Foundation Models makes Strong Few-shot Learners

Open in new window