Hard Prompts Made Easy: Gradient-Based Discrete Optimization for Prompt Tuning and Discovery

Open in new window