clarified conversation mapping

changliu2 · changliu2 · commit 0af7fc14eac8 · 2024-11-07T18:46:52.000-05:00
diff --git a/articles/ai-studio/how-to/develop/evaluate-sdk.md b/articles/ai-studio/how-to/develop/evaluate-sdk.md
@@ -125,7 +125,7 @@ A conversation is a Python dictionary of a list of messages (which include conte
 }
 ```
 
-Conversations are evaluated per turn and results are aggregated over all turns for a conversation score.
+Our evaluators will understand that the first turn of the conversation provides valid `query` from `user`, `context` from `assistant`,  and `response` from `assistant` for a single-turn evaluation in the query-response format, and that in the second turn, even if `context` is `null` or a missing key, it will be interpreted as an empty string. Conversations are then evaluated per turn and results are aggregated over all turns for a conversation score.
 
 ### Performance and quality evaluators
 

Original file line number	Diff line number	Diff line change
`@@ -125,7 +125,7 @@ A conversation is a Python dictionary of a list of messages (which include conte`
`125`	`125`	`}`
`126`	`126`	```
`127`	`127`
`128`		`-Conversations are evaluated per turn and results are aggregated over all turns for a conversation score.`
	`128`	+Our evaluators will understand that the first turn of the conversation provides valid `query` from `user`, `context` from `assistant`, and `response` from `assistant` for a single-turn evaluation in the query-response format, and that in the second turn, even if `context` is `null` or a missing key, it will be interpreted as an empty string. Conversations are then evaluated per turn and results are aggregated over all turns for a conversation score.
`129`	`129`
`130`	`130`	`### Performance and quality evaluators`
`131`	`131`