Fix: Normalize embedding model ID comparison across datasets (#5169)

Modify embedding model ID comparison to remove vendor suffixes, ensuring
consistent model identification when working with multiple knowledge
bases. This change affects dialog creation, chat operations, and
document retrieval test functions.

### What problem does this PR solve?

resolve this bug: https://github.com/infiniflow/ragflow/issues/5166

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

---------

Co-authored-by: wenju.li <wenju.li@deepctr.cn>

api/apps/dialog_app.py

@@ -18,6 +18,7 @@ from flask import request
 from flask_login import login_required, current_user
 from api.db.services.dialog_service import DialogService
 from api.db import StatusEnum
+from api.db.services.llm_service import TenantLLMService
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.user_service import TenantService, UserTenantService
 from api import settings
@@ -75,7 +76,8 @@ def set_dialog():
         if not e:
             return get_data_error_result(message="Tenant not found!")
         kbs = KnowledgebaseService.get_by_ids(req.get("kb_ids"))
-        embd_count = len(set([kb.embd_id for kb in kbs]))
+        embd_ids = [TenantLLMService.split_model_name_and_factory(kb.embd_id)[0] for kb in kbs]  # remove vendor suffix for comparison
+        embd_count = len(set(embd_ids))
         if embd_count != 1:
             return get_data_error_result(message=f'Datasets use different embedding models: {[kb.embd_id for kb in kbs]}"')