An optimized deep prompt tuning strategy comparable to fine-tuning across scales and tasks
Primary LanguagePythonApache License 2.0Apache-2.0