Revert "RAG: silence external prefetch model when chunks off-topic"
This reverts commit 1eab078311.
This commit is contained in:
parent
1eab078311
commit
538b69cf00
1 changed files with 2 additions and 6 deletions
|
|
@ -1809,12 +1809,8 @@ export function createOpenAIStreamAdapter(): ChatModelAdapter {
|
|||
const chunkXml = formatRagChunksXml(result.hits);
|
||||
const injected =
|
||||
"Use the document excerpts below to answer the question, and " +
|
||||
"cite each excerpt you use as `[N]` using its `id`. " +
|
||||
"If they are not relevant to the question, IGNORE them " +
|
||||
"silently and answer normally — do NOT mention the " +
|
||||
"excerpts, do NOT say they are irrelevant, do NOT refer to " +
|
||||
"any search or retrieval. Just answer the question " +
|
||||
"directly as if no excerpts were provided.\n\n" +
|
||||
"cite each excerpt you use as `[N]` using its `id`. If they " +
|
||||
"are not relevant, answer normally.\n\n" +
|
||||
chunkXml;
|
||||
// Send-only mutation: the displayed user bubble comes from the
|
||||
// runtime message store, not outboundMessages, so the injected
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue