Validation error in building simple Langgraph with PythonREPL Tool - Need Help

[mraguth]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

Reference: https://github.com/langchain-ai/langgraph/blob/73d1051034cd03a2de9a9df29f377243050f290a/docs/docs/tutorials/introduction.ipynb

Code I am getting error: 
### Add Memory
from langgraph.checkpoint.memory import MemorySaver
memory = MemorySaver()  

from typing import Annotated

from langchain_community.tools.tavily_search import TavilySearchResults
from langchain_core.messages import BaseMessage
from typing_extensions import TypedDict

from langgraph.graph import StateGraph, START, END
from langgraph.graph.message import add_messages
from langgraph.prebuilt import ToolNode, tools_condition


class State(TypedDict):
    messages: Annotated[list, add_messages]


graph_builder = StateGraph(State)

########################################################################## Tool #################################
class PythonInputs(BaseModel):
    query: str = Field(description="code snippet to run")

repl = PythonAstREPLTool(
    locals={"df": df},
    name="python_repl",
    description="Runs code and returns the output of the final line",
    args_schema=PythonInputs,
)

tools_repl = [repl] 

tools = [tools_repl]
llm = llm
llm_with_tools = llm.bind_tools(tools)

######################################################################################################################

def chatbot(state: State):
    return {"messages": [llm_with_tools.invoke(state["messages"])]}


graph_builder.add_node("chatbot", chatbot)

tool_node = ToolNode(tools=[tool])
graph_builder.add_node("tools", tool_node)

graph_builder.add_conditional_edges(
    "chatbot",
    tools_condition,
)
# Any time a tool is called, we return to the chatbot to decide the next step
graph_builder.add_edge("tools", "chatbot")
graph_builder.add_edge(START, "chatbot")

Description

While I am running above code, I am getting error:

{
	"name": "ValidationError",
	"message": "1 validation error for PythonAstREPLTool
args_schema
  subclass of BaseModel expected (type=type_error.subclass; expected_class=BaseModel)",
	"stack": "---------------------------------------------------------------------------
ValidationError                           Traceback (most recent call last)
Cell In[25], line 26
     23 class PythonInputs(BaseModel):
     24     query: str = Field(description=\"code snippet to run\")
---> 26 repl = PythonAstREPLTool(
     27     locals={\"df\": df},
     28     name=\"python_repl\",
     29     description=\"Runs code and returns the output of the final line\",
     30     args_schema=PythonInputs,
     31 )
     33 tools_repl = [repl] 
     35 tools = [tools_repl]

File c:\\ProgramData\\anaconda3\\envs\\langchain\\Lib\\site-packages\\langchain_core\\tools\\base.py:346, in BaseTool.__init__(self, **kwargs)
    341     if not is_basemodel_subclass(kwargs[\"args_schema\"]):
    342         raise TypeError(
    343             f\"args_schema must be a subclass of pydantic BaseModel. \"
    344             f\"Got: {kwargs['args_schema']}.\"
    345         )
--> 346 super().__init__(**kwargs)

File c:\\ProgramData\\anaconda3\\envs\\langchain\\Lib\\site-packages\\langchain_core\\load\\serializable.py:113, in Serializable.__init__(self, *args, **kwargs)
    111 def __init__(self, *args: Any, **kwargs: Any) -> None:
    112     \"\"\"\"\"\"
--> 113     super().__init__(*args, **kwargs)

File c:\\ProgramData\\anaconda3\\envs\\langchain\\Lib\\site-packages\\pydantic\\v1\\main.py:341, in BaseModel.__init__(__pydantic_self__, **data)
    339 values, fields_set, validation_error = validate_model(__pydantic_self__.__class__, data)
    340 if validation_error:
--> 341     raise validation_error
    342 try:
    343     object_setattr(__pydantic_self__, '__dict__', values)

ValidationError: 1 validation error for PythonAstREPLTool
args_schema
  subclass of BaseModel expected (type=type_error.subclass; expected_class=BaseModel)"
}

System Info

Name: pydantic
Version: 2.5.3

Name: langchain
Version: 0.2.16

[dosubot[bot]]

I found a similar unsolved discussion that might be relevant to your issue: langchain agent to use tool but tool return ValidationError ^[1].

To resolve the "subclass of BaseModel expected" validation error when building a simple Langgraph with the PythonREPL Tool in LangChain, ensure that the args_schema attribute in your tool class is correctly defined as a subclass of BaseModel. Here is an example of how to define it correctly:

For Pydantic v1:

from pydantic.v1 import BaseModel
from langchain_core.tools.base import BaseTool

class PythonInputs(BaseModel):
    query: str = Field(description="code snippet to run")

repl = PythonAstREPLTool(
    locals={"df": df},
    name="python_repl",
    description="Runs code and returns the output of the final line",
    args_schema=PythonInputs,
)

For Pydantic v2:

from pydantic import BaseModel
from langchain_core.tools.base import BaseTool

class PythonInputs(BaseModel):
    query: str = Field(description="code snippet to run")

repl = PythonAstREPLTool(
    locals={"df": df},
    name="python_repl",
    description="Runs code and returns the output of the final line",
    args_schema=PythonInputs,
)

In these examples, PythonInputs is a subclass of BaseModel, and it is assigned to the args_schema attribute of PythonAstREPLTool. This should resolve the validation error. Make sure to use the appropriate import for BaseModel based on the Pydantic version you are using ^[2][3].

To continue talking to Dosu, mention @dosu.

[mraguth]

@dosu No luck...seems like I need to define the function and call that as a tool instead of directly calling 'repl'... thoughts?

[dosubot[bot]]

Yes, you need to define a function and call that as a tool instead of directly calling repl to resolve the validation error in LangChain. Here is an example of how to define and use the PythonREPLTool correctly:

Define the Tool

from pydantic import BaseModel, Field
from langchain_core.tools import BaseTool
from langchain_experimental.utilities import PythonREPL

class PythonInputs(BaseModel):
    query: str = Field(description="code snippet to run")

class PythonAstREPLTool(BaseTool):
    name: str = "python_repl"
    description: str = "Runs code and returns the output of the final line"
    args_schema: Type[BaseModel] = PythonInputs
    python_repl: PythonREPL = PythonREPL()

    def _run(self, query: str) -> Any:
        return self.python_repl.run(query)

    async def _arun(self, query: str) -> Any:
        return await self.python_repl.arun(query)

Use the Tool

repl_tool = PythonAstREPLTool()

tools = [repl_tool]
llm_with_tools = llm.bind_tools(tools)

def chatbot(state: State):
    return {"messages": [llm_with_tools.invoke(state["messages"])]}

graph_builder.add_node("chatbot", chatbot)

tool_node = ToolNode(tools=[repl_tool])
graph_builder.add_node("tools", tool_node)

graph_builder.add_conditional_edges("chatbot", tools_condition)
graph_builder.add_edge("tools", "chatbot")
graph_builder.add_edge(START, "chatbot")

This approach ensures that the tool is properly validated and managed within the LangChain framework ^[1][2].

To continue talking to Dosu, mention @dosu.

[hwchase17]

how are you importing BaseModel? from pydantic or langchain_core.pydantic_v1?

[mraguth]

@hwchase17 I am using langchain_core.pydantic_v1, but I am open to using any base model. My objective is to achieve the same functionality in LangGraph, starting with simple steps such as reading a DataFrame into LangGraph, executing Python REPL functions, and building the graph. I see great potential in using pandas within the graph. Please help and give some sample code to explore.

[eyurtsev]

Please include the output of python -m langchain_core.sys_info -- usually impossible to help efficiently without it

[mraguth]

@eyurtsev Please take a look

(langchain) C:\Users\a024377>python -m langchain_core.sys_info

System Information
------------------
> OS:  Windows
> OS Version:  10.0.19045
> Python Version:  3.12.4 | packaged by Anaconda, Inc. | (main, Jun 18 2024, 15:03:56) [MSC v.1929 64 bit (AMD64)]

Package Information
-------------------
> langchain_core: 0.2.38
> langchain: 0.2.16
> langchain_community: 0.2.16
> langsmith: 0.1.118
> langchain_chroma: 0.1.3
> langchain_experimental: 0.0.65
> langchain_milvus: 0.1.4
> langchain_openai: 0.1.23
> langchain_text_splitters: 0.2.2
> langgraph: 0.2.19
> langserve: 0.2.3

Other Dependencies
------------------
> aiohttp: 3.10.5
> async-timeout: Installed. No version info available.
> chromadb: 0.5.3
> dataclasses-json: 0.6.7
> fastapi: 0.112.2
> httpx: 0.27.2
> jsonpatch: 1.33
> langgraph-checkpoint: 1.0.8
> numpy: 1.26.4
> openai: 1.42.0
> orjson: 3.10.7
> packaging: 24.1
> pydantic: 2.9.1
> pymilvus: 2.4.5
> pyproject-toml: 0.0.10
> PyYAML: 6.0.2
> requests: 2.32.3
> scipy: 1.13.1
> SQLAlchemy: 2.0.32
> sse-starlette: 2.1.3
> tenacity: 8.5.0
> tiktoken: 0.7.0
> typing-extensions: 4.12.2