Help Center/ Cognitive Engagement Center/ User Guide/ Tenant Administrator Guide/ Robot Management and Configuration Guide/ Large Language Model Service Zone/ LLM-based Application Services/ Mining the FAQ Knowledge Base Based on a Large Language Model/ Configuring Information on the LLM Engine
Updated on 2025-01-23 GMT+08:00
View PDF
Share

Configuring Information on the LLM Engine

Prerequisites

Contact the system administrator to sign in to the AICC, choose Call Center > Tenant Management, and enable the LLM Engine and FAQ-assisted mining features for the tenant space.

Procedure

  1. Sign in to the AICC as a tenant administrator.
  2. Choose Configuration Center > Large Model Service > Prompt Template. The prompt template management page is displayed.

    Figure 1 Prompt template management page

  3. Choose Preset Template > question answer template and click Copy to create a user-defined template.

    Figure 2 Prompt template details page
    Figure 3 Template creation page
    • Scene: Indicate the template application scenario. You can click to create a scenario.
    • Template Name: Indicate the prompt template name.
    • Language: Only Chinese and English are supported.
    • Version: This parameter must start with letter v or V followed by at least one digit.
    • Template Content: Generate the LLM prompt template.
    • content: Indicates the variable in the template content. The input content can be submitted for assessment.
    • Large Model: Select a large language model service created by the tenant.
    • Model Parameters: Indicate the model parameter switch.
    • Submit Review: Test the Q&A pair extraction result related to the current content.
    • Save and Publish: Save and publish the prompt template.

  4. Sign in to the AICC as a tenant administrator, choose Configuration Center > Chatbot Management > Large Model Service > Large Model Zone > Document management. Click to create a knowledge base.

    Figure 4 Document management page

  5. Choose Document Management > Documentation Enabled and click Upload to upload documents to the knowledge base.

    Figure 5 Document upload page

  6. After the document status changes to Read successfully, click the document name to view the text segments.

    Figure 6 Document details page

  7. Click Generate. Then the document status changes to Generated successfully. Q&A pairs are generated based on the text segments.
  8. Wait for the result generation to be complete. You can edit, accept, or discard the Q&A pairs.

    Figure 7 Q&A result generation page