Cutting Down on Prompts and Parameters: Simple Few-Shot Learning with Language Models

32580

Knowledge Triples