Migrate the MVP into kotaemon (#108)

- Migrate the MVP into kotaemon.
- Preliminary include the pipeline within chatbot interface.
- Organize MVP as an application.

Todo:

- Add an info panel to view the planning of agents -> Fix streaming agents' output.

Resolve: #60
Resolve: #61 
Resolve: #62

knowledgehub/agents/tools/base.py

@@ -41,7 +41,7 @@ class BaseTool(BaseComponent):
         args_schema = self.args_schema
         if isinstance(tool_input, str):
             if args_schema is not None:
-                key_ = next(iter(args_schema.__fields__.keys()))
+                key_ = next(iter(args_schema.model_fields.keys()))
                 args_schema.validate({key_: tool_input})
             return tool_input
         else:
@@ -121,9 +121,11 @@ class BaseTool(BaseComponent):
 
 
 class ComponentTool(BaseTool):
-    """
-    A Tool based on another pipeline / BaseComponent to be used
-    as its main entry point
+    """Wrapper around other BaseComponent to use it as a tool
+
+    Args:
+        component: BaseComponent-based component to wrap
+        postprocessor: Optional postprocessor for the component output
     """
 
     component: BaseComponent

knowledgehub/agents/tools/llm.py

@@ -1,13 +1,11 @@
-from typing import AnyStr, Optional, Type, Union
+from typing import AnyStr, Optional, Type
 
 from pydantic import BaseModel, Field
 
-from kotaemon.llms import LLM, AzureChatOpenAI, ChatLLM
+from kotaemon.llms import BaseLLM
 
 from .base import BaseTool, ToolException
 
-BaseLLM = Union[ChatLLM, LLM]
-
 
 class LLMArgs(BaseModel):
     query: str = Field(..., description="a search question or prompt")
@@ -21,7 +19,7 @@ class LLMTool(BaseTool):
         "are confident in solving the problem "
         "yourself. Input can be any instruction."
     )
-    llm: BaseLLM = AzureChatOpenAI.withx()
+    llm: BaseLLM
     args_schema: Optional[Type[BaseModel]] = LLMArgs
 
     def _run_tool(self, query: AnyStr) -> str: