From 095fc84cf2896c95e04888919d220ffd859a954a Mon Sep 17 00:00:00 2001 From: Kevin Hu Date: Tue, 25 Mar 2025 10:41:55 +0800 Subject: [PATCH] Fix: claude max tokens. (#6484) ### What problem does this PR solve? #6458 ### Type of change - [x] Bug Fix (non-breaking change which fixes an issue) --- rag/llm/chat_model.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/rag/llm/chat_model.py b/rag/llm/chat_model.py index ebcce26a7..1dfdb1068 100644 --- a/rag/llm/chat_model.py +++ b/rag/llm/chat_model.py @@ -1443,7 +1443,7 @@ class AnthropicChat(Base): del gen_conf["presence_penalty"] if "frequency_penalty" in gen_conf: del gen_conf["frequency_penalty"] - gen_conf["max_tokens"] = 8196 + gen_conf["max_tokens"] = 8192 if "haiku" in self.model_name or "opus" in self.model_name: gen_conf["max_tokens"] = 4096 @@ -1477,7 +1477,7 @@ class AnthropicChat(Base): del gen_conf["presence_penalty"] if "frequency_penalty" in gen_conf: del gen_conf["frequency_penalty"] - gen_conf["max_tokens"] = 8196 + gen_conf["max_tokens"] = 8192 if "haiku" in self.model_name or "opus" in self.model_name: gen_conf["max_tokens"] = 4096