refine the lamp output

KevinHuSh · Oct 11, 2024 · 9c8fa2a · 9c8fa2a
1 parent 5e7c1fb
commit 9c8fa2a
Showing 1 changed file with 4 additions and 2 deletions.
diff --git a/api/db/services/dialog_service.py b/api/db/services/dialog_service.py
@@ -153,6 +153,7 @@ def chat(dialog, messages, stream=True, **kwargs):
         questions = [full_question(dialog.tenant_id, dialog.llm_id, messages)]
     else:
         questions = questions[-1:]
+    refineQ_tm = timer()
 
     rerank_mdl = None
     if dialog.rerank_id:
@@ -197,7 +198,7 @@ def chat(dialog, messages, stream=True, **kwargs):
             max_tokens - used_token_count)
 
     def decorate_answer(answer):
-        nonlocal prompt_config, knowledges, kwargs, kbinfos, prompt, retrieval_tm
+        nonlocal prompt_config, knowledges, kwargs, kbinfos, prompt, retrieval_tm, refineQ_tm
         refs = []
         if knowledges and (prompt_config.get("quote", True) and kwargs.get("quote", True)):
             answer, idx = retr.insert_citations(answer,
@@ -222,7 +223,8 @@ def decorate_answer(answer):
         if answer.lower().find("invalid key") >= 0 or answer.lower().find("invalid api") >= 0:
             answer += " Please set LLM API-Key in 'User Setting -> Model Providers -> API-Key'"
         done_tm = timer()
-        prompt += "\n\n### Elapsed\n  - Retrieval: %.1f ms\n  - LLM: %.1f ms"%((retrieval_tm-st)*1000, (done_tm-st)*1000)
+        prompt += "\n\n### Elapsed\n  - Refine Question: %.1f ms\n  - Retrieval: %.1f ms\n  - LLM: %.1f ms" % (
+            (refineQ_tm - st) * 1000, (retrieval_tm - refineQ_tm) * 1000, (done_tm - retrieval_tm) * 1000)
         return {"answer": answer, "reference": refs, "prompt": prompt}
 
     if stream: