Question and answer model training method and device
By compressing the initial system instructions and generating training data with sample Q&A information, the Q&A model is trained, and the problem of low efficiency of the Q&A model is solved, quickly dealing with user problems and improving user experience.
CN119621913APending Publication Date: 2025-03-14BEIJING YUANLI WEILAI SCI & TECH CO LTD
View PDF 0 Cites 0 Cited by
Patent Information
- Application Number
- CN202411838162.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-13
- Publication Date
- 2025-03-14
AI Technical Summary
Technical Problem
In the prior art, the question-and-answer model is inefficient in dealing with user problems, resulting in a long wait time for the user and affecting the user experience.
Method used
By obtaining the initial system instructions and compressing according to the preset compression method, the compressed system instructions with shorter character length are obtained, the training data is generated based on the sample Q&A information, the preset language model is trained, and the question and answer model is obtained.
Benefits of technology
It reduces the data processing volume of the Q&A model, improves the processing efficiency of the Q&A model, shortens user waiting time, and improves user experience.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN119621913A_ABST
Abstract
The embodiment of the invention discloses a question and answer model training method and device. The question and answer model training method comprises the steps that an initial system instruction is acquired; the initial system instruction is a natural language description used for indicating a model to answer a question of a user in a specified role; performing compression processing on the initial system instruction according to a preset compression mode to obtain a compressed system instruction; the character length of the compressed system instruction is smaller than the character length of the initial system instruction; obtaining training data according to the compressed system instruction and sample question and answer information; the sample question and answer information comprises sample question information and answer information corresponding to the sample question information; and training a preset language model according to the training data to obtain a question and answer model.
Need to check novelty before this filing date? Find Prior Art