Large Language Models (LLMs) demonstrate impressive performance in various downstream tasks. However, they may still generate incorrect responses in certain scenarios due to the knowledge deficiencies and the flawed pre-training data. Continual Learning (CL) is a commonly used method to address this issue. Traditional CL is task-oriented, using novel or factually accurate data to retrain LLMs from scratch. However, this method requires more task-related training data and incurs expensive training costs. To address this challenge, we propose the Continue Evolving from Mistakes (CEM) method, inspired by the 'summarize mistakes' learning skill, to achieve iterative refinement of LLMs. Specifically, the incorrect responses of LLMs indicate knowledge deficiencies related to the questions. Therefore, we collect corpora with these knowledge from multiple data sources and follow it up with iterative supplementary training for continuous, targeted knowledge updating and supplementation. Meanwhile, we developed two strategies to construct supplementary training sets to enhance the LLM's understanding of the corpus and prevent catastrophic forgetting. We conducted extensive experiments to validate the effectiveness of this CL method. In the best case, our method resulted in a 17.00\% improvement in the accuracy of the LLM.

翻译：大语言模型（LLMs）在各类下游任务中展现出惊人性能，但由于知识缺陷和预训练数据存在的瑕疵，它们在特定场景下仍可能产生错误响应。持续学习（CL）是解决该问题的常用方法。传统持续学习以任务为导向，通过使用新颖或事实准确的数据从头重训练LLMs。然而，这种方法需要更多任务相关训练数据，且训练成本高昂。为应对这一挑战，我们受"总结错误"学习策略启发，提出"从错误中持续进化"（CEM）方法，以实现LLMs的迭代优化。具体而言，LLMs的错误响应揭示了其与问题相关的知识缺陷。因此，我们从多个数据源收集包含这些知识的语料库，随后通过迭代补充训练进行持续、定向的知识更新与补充。同时，我们开发了两种构建补充训练集的策略，以增强LLM对语料的理解并防止灾难性遗忘。我们进行了大量实验验证该持续学习方法的有效性。在最佳情况下，我们的方法使LLM的准确率提升17.00%。