Add as_server command and add server_id

docs/sphinx_doc/en/source/index.rst

@@ -38,6 +38,7 @@ AgentScope Documentation
    agentscope.pipelines
    agentscope.service
    agentscope.rpc
+   agentscope.server
    agentscope.web
    agentscope.prompt
    agentscope.utils

docs/sphinx_doc/en/source/tutorial/208-distribute.md

@@ -59,14 +59,16 @@ b = AgentB(
 #### Independent Process Mode
 
 In the Independent Process Mode, we need to start the agent server process on the target machine first.
+When starting the agent server process, you need to specify a model config file, which contains the models which can be used in the agent server, the IP address and port of the agent server process
 For example, start two agent server processes on the two different machines with IP `ip_a` and `ip_b`(called `Machine1` and `Machine2` accrodingly).
-You can run the following code on `Machine1`:
+You can run the following code on `Machine1`, and make sure you have put your model config file in `model_config_path_a`. The example model config file instances are located under `examples/model_configs_template`.
 
 ```python
 # import some packages
 
+# register models which can be used in the server
 agentscope.init(
-    ...
+    model_configs=model_config_path_a,
 )
 # Create an agent service process
 server = RpcAgentServerLauncher(
@@ -79,13 +81,20 @@ server.launch()
 server.wait_until_terminate()
 ```
 
-And run the following code on `Machine2`:
+> For similarity, you can run the following command in your terminal rather than the above code:
+>
+> ```shell
+> as_server --host ip_a --port 12001 --model-config-path model_config_path_a
+> ```
+
+And put your model config file accordingly in `model_config_path_b` and run the following code on `Machine2`.
 
 ```python
 # import some packages
 
+# register models which can be used in the server
 agentscope.init(
-    ...
+    model_configs=model_config_path_b,
 )
 # Create an agent service process
 server = RpcAgentServerLauncher(
@@ -98,6 +107,12 @@ server.launch()
 server.wait_until_terminate()
 ```
 
+> Similarly, you can run the following command in your terminal to setup the agent server:
+>
+> ```shell
+> as_server --host ip_b --port 12002 --model-config-path model_config_path_b
+> ```
+
 Then, you can connect to the agent servers from the main process with the following code.
 
 ```python
@@ -254,6 +269,9 @@ About more detailed technical implementation solutions, please refer to our [pap
 
 In agentscope, the agent server provides a running platform for various types of agents.
 Multiple agents can run in the same agent server and hold independent memory and other local states but they will share the same computation resources.
+
+After installing the distributed version of AgentScope, you can use the `as_server` command to start the agent server, and the detailed startup arguments can be found in the documentation of the {func}`as_server<agentscope.server.launcher.as_server>` function.
+
 As long as the code is not modified, an agent server can provide services for multiple main processes.
 This means that when running mutliple applications, you only need to start the agent server for the first time, and it can be reused subsequently.
 

docs/sphinx_doc/zh_CN/source/index.rst

@@ -38,6 +38,7 @@ AgentScope 文档
    agentscope.pipelines
    agentscope.service
    agentscope.rpc
+   agentscope.server
    agentscope.web
    agentscope.prompt
    agentscope.utils

docs/sphinx_doc/zh_CN/source/tutorial/208-distribute.md

@@ -57,15 +57,16 @@ b = AgentB(
 
 #### 独立进程模式
 
-在独立进程模式中，需要首先在目标机器上启动智能体服务器进程。
+在独立进程模式中，需要首先在目标机器上启动智能体服务器进程，启动时需要提供该服务器能够使用的模型的配置信息，以及服务器的 IP 和端口号。
 例如想要将两个智能体服务进程部署在 IP 分别为 `ip_a` 和 `ip_b` 的机器上（假设这两台机器分别为`Machine1` 和 `Machine2`）。
-你可以先在 `Machine1` 上运行如下代码：
+你可以先在 `Machine1` 上运行如下代码，运行之前请确保已经将模型配置文件放置在 `model_config_path_a` 位置，模型配置文件样例可参考 `examples/model_configs_template`。
 
 ```python
 # import some packages
 
+# register models which can be used in the server
 agentscope.init(
-    ...
+    model_configs=model_config_path_a,
 )
 # Create an agent service process
 server = RpcAgentServerLauncher(
@@ -78,13 +79,20 @@ server.launch()
 server.wait_until_terminate()
 ```
 
-之后在 `Machine2` 上运行如下代码：
+> 为了进一步简化使用，可以在命令行中输入如下指令来代替上述代码：
+>
+> ```shell
+> as_server --host ip_a --port 12001  --model-config-path model_config_path_a
+> ```
+
+在 `Machine2` 上运行如下代码，这里同样要确保已经将模型配置文件放置在 `model_config_path_b` 位置。
 
 ```python
 # import some packages
 
+# register models which can be used in the server
 agentscope.init(
-    ...
+    model_configs=model_config_path_b,
 )
 # Create an agent service process
 server = RpcAgentServerLauncher(
@@ -97,6 +105,12 @@ server.launch()
 server.wait_until_terminate()
 ```
 
+> 这里也同样可以用如下指令来代替上面的代码。
+>
+> ```shell
+> as_server --host ip_b --port 12002 --model-config-path model_config_path_b
+> ```
+
 接下来，就可以使用如下代码从主进程中连接这两个智能体服务器进程。
 
 ```python
@@ -251,6 +265,9 @@ Placeholder 内部包含了该消息产生方的联络方法，可以通过网
 #### Agent Server
 
 Agent Server 也就是智能体服务器。在 AgentScope 中，Agent Server 提供了一个让不同 Agent 实例运行的平台。多个不同类型的 Agent 可以运行在同一个 Agent Server 中并保持独立的记忆以及其他本地状态信息，但是他们将共享同一份计算资源。
+
+在安装 AgentScope 的分布式版本后就可以通过 `as_server` 命令来启动 Agent Server，具体的启动参数在 {func}`as_server<agentscope.server.launcher.as_server>` 函数文档中可以找到。
+
 只要没有对代码进行修改，一个已经启动的 Agent Server 可以为多个主流程提供服务。
 这意味着在运行多个应用时，只需要在第一次运行前启动 Agent Server，后续这些 Agent Server 进程就可以持续复用。
 

examples/distributed_basic/distributed_dialog.py

@@ -7,7 +7,7 @@
 import agentscope
 from agentscope.agents.user_agent import UserAgent
 from agentscope.agents.dialog_agent import DialogAgent
-from agentscope.agents.rpc_agent import RpcAgentServerLauncher
+from agentscope.server import RpcAgentServerLauncher
 
 
 def parse_args() -> argparse.Namespace:

examples/distributed_debate/distributed_debate.py

@@ -8,7 +8,7 @@
 import agentscope
 from agentscope.agents import DialogAgent
 from agentscope.msghub import msghub
-from agentscope.agents.rpc_agent import RpcAgentServerLauncher
+from agentscope.server import RpcAgentServerLauncher
 from agentscope.message import Msg
 from agentscope.utils.logging_utils import logger
 

examples/distributed_simulation/main.py

@@ -11,7 +11,7 @@
 
 import agentscope
 from agentscope.agents import AgentBase
-from agentscope.agents.rpc_agent import RpcAgentServerLauncher
+from agentscope.server import RpcAgentServerLauncher
 from agentscope.message import Msg
 
 

setup.py

@@ -122,6 +122,7 @@
         "console_scripts": [
             "as_studio=agentscope.web.studio.studio:run_app",
             "as_workflow=agentscope.web.workstation.workflow:main",
+            "as_server=agentscope.server.launcher:as_server",
         ],
     },
 )

src/agentscope/agents/__init__.py

@@ -6,7 +6,7 @@
 from .dict_dialog_agent import DictDialogAgent
 from .user_agent import UserAgent
 from .text_to_image_agent import TextToImageAgent
-from .rpc_agent import RpcAgent, RpcAgentServerLauncher
+from .rpc_agent import RpcAgent
 from .react_agent import ReActAgent
 
 
@@ -20,5 +20,4 @@
     "ReActAgent",
     "DistConf",
     "RpcAgent",
-    "RpcAgentServerLauncher",
 ]

src/agentscope/agents/agent.py

@@ -24,7 +24,7 @@ class _AgentMeta(ABCMeta):
     """
 
     def __init__(cls, name: Any, bases: Any, attrs: Any) -> None:
-        if not hasattr(cls, "registry"):
+        if not hasattr(cls, "_registry"):
             cls._registry = {}
         else:
             if name in cls._registry:
@@ -245,7 +245,7 @@ def register_agent_class(cls, agent_class: Type[AgentBase]) -> None:
         """
         agent_class_name = agent_class.__name__
         if agent_class_name in cls._registry:
-            logger.warning(
+            logger.info(
                 f"Agent class with name [{agent_class_name}] already exists.",
             )
         else: