From fa682ec38f77399ead94bf344e7e4ea01558ebcd Mon Sep 17 00:00:00 2001
From: SengokuCola <1026294844@qq.com>
Date: Sat, 25 Oct 2025 15:59:11 +0800
Subject: [PATCH] =?UTF-8?q?fix=EF=BC=9A=E7=A7=BB=E9=99=A4vlm=E7=9A=84?=
 =?UTF-8?q?=E7=94=9F=E6=88=90=E9=99=90=E5=88=B6?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 src/chat/utils/utils_image.py       | 4 ++--
 template/model_config_template.toml | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/src/chat/utils/utils_image.py b/src/chat/utils/utils_image.py
index deba60a4..601373a1 100644
--- a/src/chat/utils/utils_image.py
+++ b/src/chat/utils/utils_image.py
@@ -273,7 +273,7 @@ class ImageManager:
             prompt = global_config.personality.visual_style
             logger.info(f"[VLM调用] 为图片生成新描述 (Hash: {image_hash[:8]}...)")
             description, _ = await self.vlm.generate_response_for_image(
-                prompt, image_base64, image_format, temperature=0.4, max_tokens=300
+                prompt, image_base64, image_format, temperature=0.4
             )
 
             if description is None:
@@ -570,7 +570,7 @@ class ImageManager:
 
             # 获取VLM描述
             description, _ = await self.vlm.generate_response_for_image(
-                prompt, image_base64, image_format, temperature=0.4, max_tokens=300
+                prompt, image_base64, image_format, temperature=0.4
             )
 
             if description is None:
diff --git a/template/model_config_template.toml b/template/model_config_template.toml
index c97a6579..9f2b05d5 100644
--- a/template/model_config_template.toml
+++ b/template/model_config_template.toml
@@ -1,5 +1,5 @@
 [inner]
-version = "1.7.4"
+version = "1.7.5"
 
 # 配置文件版本号迭代规则同bot_config.toml
 
@@ -120,7 +120,7 @@ max_tokens = 800
 
 [model_task_config.vlm] # 图像识别模型
 model_list = ["qwen2.5-vl-72b"]
-max_tokens = 800
+max_tokens = 256
 
 [model_task_config.voice] # 语音识别模型
 model_list = ["sensevoice-small"]