Merge pull request modelscope#18 from ZiTao-Li/zitao/dev_copilot

update as comments

docs/sphinx_doc/zh_CN/source/tutorial/209-rag.md

@@ -5,7 +5,7 @@
 我们在此介绍AgentScope与RAG相关的三个概念：知识（Knowledge），知识库（Knowledge Bank）和RAG 智能体。
 
 ### Knowledge
-知识模块（目前仅有“LlamaIndexKnowledge”；即将支持对LangChain）负责处理所有与RAG相关的操作。
+知识模块（目前仅有“LlamaIndexKnowledge”；即将提供对LangChain的支持）负责处理所有与RAG相关的操作。
 
 在这里，我们将使用`LlamaIndexKnowledge`作为示例，以说明在`Knowledge`模块内的操作。
 当初始化`LlamaIndexKnowledge`对象时，`LlamaIndexKnowledge.__init__`将执行以下步骤：

src/agentscope/agents/rag_agents.py

@@ -13,7 +13,6 @@
 from agentscope.message import Msg
 from agentscope.rag import Knowledge
 
-
 CHECKING_PROMPT = """
                 Is the retrieved content relevant to the query?
                 Retrieved content: {}
@@ -22,6 +21,29 @@
                 """
 
 
+class RAGConfig(dict):
+    """a class to regulate the RAG configuration."""
+
+    def __init__(self, knowledge_id: list[str], **kwargs: Any):
+        """
+        RAG configuration must have knowledge_id as a list of strings.
+        Args:
+            knowledge_id (list[str]): the list of knowledge ids
+        """
+        super().__init__()
+        self.knowledge_id = knowledge_id
+        self.update(kwargs)
+
+    def __getattr__(self, key: Any) -> Any:
+        try:
+            return self[key]
+        except KeyError as e:
+            raise AttributeError(f"no attribute '{key}'") from e
+
+    def __setattr__(self, key: Any, value: Any) -> None:
+        self[key] = value
+
+
 class LlamaIndexAgent(AgentBase):
     """
     A LlamaIndex agent build on LlamaIndex.
@@ -34,7 +56,7 @@ def __init__(
         model_config_name: str,
         knowledge_list: list[Knowledge] = None,
         memory_config: Optional[dict] = None,
-        rag_config: Optional[dict] = None,
+        rag_config: Optional[RAGConfig] = None,
         **kwargs: Any,
     ) -> None:
         """
@@ -70,7 +92,7 @@ def __init__(
         self.knowledge_list = knowledge_list or []
         self.retriever_list = []
         self.description = kwargs.get("description", "")
-        self.rag_config = rag_config or {}
+        self.rag_config = rag_config or RAGConfig([])
 
     def reply(self, x: dict = None) -> dict:
         """
@@ -143,7 +165,8 @@ def reply(self, x: dict = None) -> dict:
                         query,
                     ),
                 )
-                checking = self.model([msg])
+                msg = self.model.format(msg)
+                checking = self.model(msg)
                 logger.info(checking)
                 checking = checking.text.lower()
                 if "no" in checking:

src/agentscope/constants.py

@@ -59,3 +59,9 @@ class ShrinkPolicy(IntEnum):
 
     TRUNCATE = 0
     SUMMARIZE = 1
+
+
+# rag related
+DEFAULT_CHUNK_SIZE = 1024
+DEFAULT_CHUNK_OVERLAP = 20
+DEFAULT_TOP_K = 5

src/agentscope/rag/knowledge.py

@@ -18,10 +18,6 @@
 from loguru import logger
 from agentscope.models import ModelWrapperBase
 
-DEFAULT_CHUNK_SIZE = 1024
-DEFAULT_CHUNK_OVERLAP = 20
-DEFAULT_TOP_K = 5
-
 
 class Knowledge(ABC):
     """

src/agentscope/rag/knowledge_bank.py

@@ -163,6 +163,7 @@ def equip(self, agent: AgentBase, duplicate: bool = False) -> None:
             agent (AgentBase): the agent to be equipped with knowledge
                 if it has "rag_config"
             duplicate (bool): whether to deepcopy the knowledge object
+        TODO: to accommodate with distributed setting
         """
         if hasattr(agent, "rag_config") and "knowledge_id" in agent.rag_config:
             if not hasattr(agent, "knowledge_list"):

src/agentscope/rag/llama_index_knowledge.py

@@ -27,24 +27,29 @@
         Document,
         TransformComponent,
     )
-except ImportError:
-    BaseRetriever = None
-    BaseEmbedding, Embedding = None, None
-    IngestionPipeline = None
-    SentenceSplitter = None
-    VectorStoreIndex, StorageContext = None, None
-    load_index_from_storage = None
-    PrivateAttr = None
-    Document, TransformComponent = None, None
+except ImportError as import_error:
+    from agentscope.utils.tools import ImportErrorReporter
+
+    BaseRetriever = ImportErrorReporter(import_error, "llama-index")
+    BaseEmbedding = ImportErrorReporter(import_error, "llama-index")
+    Embedding = ImportErrorReporter(import_error, "llama-index")
+    IngestionPipeline = ImportErrorReporter(import_error, "llama-index")
+    SentenceSplitter = ImportErrorReporter(import_error, "llama-index")
+    VectorStoreIndex = ImportErrorReporter(import_error, "llama-index")
+    StorageContext = ImportErrorReporter(import_error, "llama-index")
+    load_index_from_storage = ImportErrorReporter(import_error, "llama-index")
+    PrivateAttr = ImportErrorReporter(import_error, "llama-index")
+    Document = ImportErrorReporter(import_error, "llama-index")
+    TransformComponent = ImportErrorReporter(import_error, "llama-index")
 
 from agentscope.file_manager import file_manager
 from agentscope.models import ModelWrapperBase
-from .knowledge import (
+from agentscope.constants import (
     DEFAULT_TOP_K,
     DEFAULT_CHUNK_SIZE,
     DEFAULT_CHUNK_OVERLAP,
-    Knowledge,
 )
+from agentscope.rag.knowledge import Knowledge
 
 
 try:
@@ -426,7 +431,8 @@ def set_retriever(
         Set the retriever as needed, or just use the default setting.
 
         Args:
-            retriever (Optional[BaseRetriever]): passing a retriever in llama
+            retriever (Optional[BaseRetriever]): passing a retriever in
+             LlamaIndexKnowledge
             rag_config (dict): rag configuration, including similarity top k
             index.
         """