fix: defer missing-local-embedding error so it can't crash boot (#25683)

get_embedding_function runs at import time (main.py), so the empty-engine
+ no-model guard added in 55ca719b raised ValueError during startup and
bricked the instance: a blank embedding model set via the UI made the app
unbootable, with no way back into settings to undo it. Move the check into
the returned coroutine so construction always succeeds and the error only
fires when something actually embeds, surfacing as a clear request error
instead of the old cryptic 'NoneType has no attribute encode'.


Fixes #25634
Fixes #25165

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

backend/open_webui/retrieval/utils.py

@@ -1089,15 +1089,15 @@ def get_embedding_function(
     concurrent_requests=0,
 ) -> Awaitable:
     if embedding_engine == '':
-        if embedding_function is None:
-            raise ValueError(
-                'No embedding model is loaded. Set RAG_EMBEDDING_MODEL to a valid '
-                'SentenceTransformer model name, or configure an external '
-                'RAG_EMBEDDING_ENGINE (ollama, openai, azure_openai).'
-            )
-
         # Sentence transformers: CPU-bound sync operation
         async def async_embedding_function(query, prefix=None, user=None):
+            # Deferred so a missing local model degrades RAG instead of crashing boot.
+            if embedding_function is None:
+                raise ValueError(
+                    'No embedding model is loaded. Set RAG_EMBEDDING_MODEL to a valid '
+                    'SentenceTransformer model name, or configure an external '
+                    'RAG_EMBEDDING_ENGINE (ollama, openai, azure_openai).'
+                )
             return await asyncio.to_thread(
                 (
                     lambda query, prefix=None: embedding_function.encode(