Clarification on variables used in orchestration prompt

Question

The following is a small snippet of the orchestration prompt, more specifically the '$additional_kb_guidelines$' prompt.

Structure any answer that uses information from search results in this 
      exact XML format:

<source>$SOURCE_ID$

I want to know what is the $ANSWER_PART$ and $SOURCE_ID$ refer to.

Devraj_P · Accepted Answer

Hello,

I understand that you are using Bedrock and would like clarification on the variables used in the orchestration prompt, specifically $ANSWER_PART$ and $SOURCE_ID$.

**$ANSWER_PART$** – This is where the actual text content of the answer goes. It's the substantive information being provided in response to the query.

**$SOURCE_ID$** – Refers to the unique identifier of the source document or reference from which the information in the answer part was obtained. This helps track and validate the origin of the information.
 
```
Example -
If you were answering a question about climate change using search results. 
$ANSWER_PART$ would be replaced with "Global temperatures have risen by approximately 1.1°C since pre-industrial times."
$SOURCE_ID$ would be replaced with "IPCC Climate Report 2023"
```
If you encounter any difficulties, please reach out to AWS Support [1] with details of your use case, and we will be happy to assist you further.

—References—

[1] Creating support cases and case management: https://docs.aws.amazon.com/awssupport/latest/user/case-management.html#creating-a-support-case

Clarification on variables used in orchestration prompt

Füge deine Antwort hinzu

Relevanter Inhalt