diff --git a/Pasted image 20250317180130.png b/Pasted image 20250317180130.png
new file mode 100644
index 0000000..d482c35
Binary files /dev/null and b/Pasted image 20250317180130.png differ
diff --git a/Pasted image 20250317180245.png b/Pasted image 20250317180245.png
new file mode 100644
index 0000000..d70a517
Binary files /dev/null and b/Pasted image 20250317180245.png differ
diff --git a/日常学习/大模型/function calling.md b/日常学习/大模型/function calling.md
new file mode 100644
index 0000000..a4796c2
--- /dev/null
+++ b/日常学习/大模型/function calling.md	
@@ -0,0 +1,58 @@
+# 一、概念
+function Calling 是一种机制，允许大语言模型[动态调用](https://zhida.zhihu.com/search?content_id=254823414&content_type=Article&match_order=1&q=%E5%8A%A8%E6%80%81%E8%B0%83%E7%94%A8&zd_token=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJ6aGlkYV9zZXJ2ZXIiLCJleHAiOjE3NDIzNzc4MTgsInEiOiLliqjmgIHosIPnlKgiLCJ6aGlkYV9zb3VyY2UiOiJlbnRpdHkiLCJjb250ZW50X2lkIjoyNTQ4MjM0MTQsImNvbnRlbnRfdHlwZSI6IkFydGljbGUiLCJtYXRjaF9vcmRlciI6MSwiemRfdG9rZW4iOm51bGx9.SFQ4VSfJ2VWn-0GlIyT2_oZcB3dIPRDbBWFy9g7gYmM&zhida_source=entity)外部函数或API，以完成特定任务。例如，当用户提问“今天的天气如何？”时，模型可以通过调用天气API获取实时数据并返回结果。这种方式不仅提升了模型的实用性，还使其能够处理复杂的多步骤任务。
+
+流程如下：
+![[Pasted image 20250317180245.png]]
+
+1. 用户输入问题或指令。
+2. 模型分析输入，生成 [JSON 格式](https://zhida.zhihu.com/search?content_id=254823414&content_type=Article&match_order=1&q=JSON+%E6%A0%BC%E5%BC%8F&zd_token=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJ6aGlkYV9zZXJ2ZXIiLCJleHAiOjE3NDIzNzc4MTgsInEiOiJKU09OIOagvOW8jyIsInpoaWRhX3NvdXJjZSI6ImVudGl0eSIsImNvbnRlbnRfaWQiOjI1NDgyMzQxNCwiY29udGVudF90eXBlIjoiQXJ0aWNsZSIsIm1hdGNoX29yZGVyIjoxLCJ6ZF90b2tlbiI6bnVsbH0.CqcXsaw3SWw2cR64Zo3MC3hEnVXL7SwKfAfkt-xau9I&zhida_source=entity)的函数调用请求。
+3. 后端解析请求，调用对应函数或API。
+4. 将函数返回的结果传递回模型。
+5. 模型基于结果生成最终回答。
+
+# 二、Deepseek function calling
+
+1. tools定义
+deepseek 官方定义的格式（最多支持 128 个 function），参考 https://api-docs.deepseek.com/zh-cn/guides/function_calling
+```json
+{  
+    "type": "function",  
+    "function": {  
+        "name": "get_current_weather",  
+        "description": "Get weather of an location, the user shoud supply a location first",  
+        "parameters": {  
+            "type": "object",  
+            "properties": {  
+                "location": {  
+                    "type": "string",  
+                    "description": "The city and state, e.g. San Francisco, CA",  
+                }  
+            },  
+            "required": ["location"]  
+        },  
+    }  
+}
+```
+
+2. 函数绑定及调用
+```python
+模型初始化
+llm = ChatOpenAI(  
+    model='deepseek-r1-250120', api_key='4eefc827-187f-4756-9637-7e0153c93d81',  
+    base_url='https://ark.cn-beijing.volces.com/api/v3/', max_tokens=4096, temperature=0.5, streaming=True  
+)
+
+
+@staticmethod  
+def invoke_with_tool(user_input: str):  
+    """  
+    工具链调用,function calling时system prompt不会生效  
+    :param user_input:    :return: 这里返回的是LLM推理出的tool信息，格式如下：  
+        [{'name': 'get_current_weather', 'args': {'location': 'Beijing, China'}, 'id': 'call_xeeq4q52fw9x61lkrqwy9cr6', 'type': 'tool_call'}]    """    llm_with_tools = llm.bind_tools(TOOLS)  
+    return llm_with_tools.invoke(user_input).tool_calls
+```
+
+3. 返回样例,包含函数名和参数
+```json
+[{'name': 'get_current_weather', 'args': {'location': 'Beijing, China'}, 'id': 'call_xeeq4q52fw9x61lkrqwy9cr6', 'type': 'tool_call'}]
+```
diff --git a/本机环境/开发环境.md b/本机环境/开发环境.md
index 2a902f0..338a62d 100644
--- a/本机环境/开发环境.md
+++ b/本机环境/开发环境.md
@@ -208,3 +208,9 @@ cd /opt/homebrew/Cellar/apache-flink/1.20.1/libexec/bin && ./start-cluster.sh
 ```
 web UI
 http://localhost:8081/#/overview
+
++ marqo
+```shell
+docker pull marqoai/marqo:latest
+docker run --name marqo -it --privileged -p 8882:8882 --add-host host.docker.internal:host-gateway marqoai/marqo:latest
+```