ConceptCPT / Continual Pre-Training
继续预训练
在通用基座模型上用领域语料进行额外无监督训练，使模型吸收领域知识的技术方法，属于迁移学习的一种形式
Core Facts
Timeline (last 90 days)
Oct 12
Learning-based methods have stronger performance when data is sufficient, but at the cost of robustness under domain shift.
Unverified50%
All Facts (2)
Verified
Fine-tuning continues to update parameters on top of pre-trained weights through the backpropagation algorithm, adapting the model to a specific domain or task style.
75%UnverifiedLearning-based methods have stronger performance when data is sufficient, but at the cost of robustness under domain shift.
50%