zero-shot
Zero-shot is the natural best-case scenario for a model as it means we require zero training samples before shifting it to a new domain or task.
zero-shot是一个模型的自然最佳情况,因为这意味着我们将其迁移到新的领域或者任务时需要零个训练样本
Zero-shot is the natural best-case scenario for a model as it means we require zero training samples before shifting it to a new domain or task.
zero-shot是一个模型的自然最佳情况,因为这意味着我们将其迁移到新的领域或者任务时需要零个训练样本