Translations:FACTS About Building Retrieval Augmented Generation-based Chatbots/53/zh: Difference between revisions

Latest revision as of 08:52, 19 February 2025

Information about message (contribute)

This message has no documentation. If you know where or how this message is used, you can help other translators by adding documentation to this message.

Message definition (FACTS About Building Retrieval Augmented Generation-based Chatbots)

* '''Feedback Loops:''' Incorporating feedback gathered and the RLHF cycle is pivotal for continuous improvement. It allows LLM models to refine both our solutions and Language Models over time, ensuring that the chatbot becomes increasingly proficient. However, if the chosen foundational models don’t offer customization, then it becomes difficult to align the models to human feedback. If the feedback is significant and comes in many areas, then model customization may be an option. As of now, we have begun gathering user feedback but haven’t built our continuous learning pipelines using RLHF yet. Having tools to make this automated is critical to post-production life cycle management of these chatbots.

反馈循环： 结合收集到的反馈和RLHF循环对于持续改进至关重要。它使得LLM模型能够随着时间的推移不断优化我们的解决方案和语言模型，确保聊天机器人变得越来越熟练。然而，如果所选择的基础模型不提供定制化，那么就很难将模型与人类反馈对齐。如果反馈在多个领域中显著且广泛，那么模型定制可能是一个选项。目前，我们已经开始收集用户反馈，但尚未使用RLHF构建我们的持续学习管道。拥有能够实现自动化的工具对于这些聊天机器人的后期生产生命周期管理至关重要。