mlflow/mlflow · warning · ValueError

Invalid data for {pydantic_class.__name__}: {e}

Error message

Invalid data for {pydantic_class.__name__}: {e}

What it means

validate_pydantic attempts to construct pydantic_class from the data dict (or re-validate a BaseModel via model_dump). Any construction/validation failure is re-raised as ValueError naming the model class, which the server surfaces as a 400 error.

Source

Thrown at mlflow/genai/agent_server/validator.py:26

    ResponsesAgentResponse,
    ResponsesAgentStreamEvent,
)


class BaseAgentValidator:
    """Base validator class with common validation methods"""

    def validate_pydantic(self, pydantic_class: type[BaseModel], data: Any) -> None:
        """Generic pydantic validator that throws an error if the data is invalid"""
        if isinstance(data, pydantic_class):
            return
        try:
            if isinstance(data, BaseModel):
                pydantic_class(**data.model_dump())
                return
            pydantic_class(**data)
        except Exception as e:
            raise ValueError(f"Invalid data for {pydantic_class.__name__}: {e}")

    def validate_dataclass(self, dataclass_class: Any, data: Any) -> None:
        """Generic dataclass validator that throws an error if the data is invalid"""
        if isinstance(data, dataclass_class):
            return
        try:
            dataclass_class(**data)
        except Exception as e:
            raise ValueError(f"Invalid data for {dataclass_class.__name__}: {e}")

    def validate_and_convert_request(self, data: dict[str, Any]) -> dict[str, Any]:
        return data

    def validate_and_convert_result(self, result: Any, stream: bool = False) -> dict[str, Any]:
        # Base implementation doesn't use stream parameter, but subclasses do
        if isinstance(result, BaseModel):
            return result.model_dump(exclude_none=True)
        elif is_dataclass(result):

View on GitHub (pinned to 6a27f2decc)

Solutions

1. Read the wrapped pydantic error (included in the message) and fix the offending field
2. Validate your payload locally with pydantic_class(**data) before calling the endpoint
3. Regenerate payloads from the current SDK/model schema rather than hand-building dicts

Example fix

# before
validate_pydantic(ChatAgentRequest, {"messages": "hi"})  # messages must be a list

# after
validate_pydantic(ChatAgentRequest, {"messages": [{"role": "user", "content": "hi"}]})

try:
    PydanticClass(**payload)
except Exception as e:
    raise ValueError(f"payload invalid before send: {e}")

from pydantic import BaseModel
from typing import Type, Any

def matches_model(model: Type[BaseModel], data: Any) -> bool:
    if not isinstance(data, dict):
        return False
    try:
        model(**data)
        return True
    except Exception:
        return False

try:
    result = client.predict(payload)
except ValueError as e:
    if "Invalid data for" in str(e):
        logger.error("Fix fields per pydantic error: %s", e)

When it happens

Trigger: Passing a dict missing required pydantic fields, with wrong field types, or a BaseModel instance whose dumped fields no longer satisfy the target class.

Common situations: Client sending payloads that don't match the agent's input/output pydantic models; schema drift between SDK versions; nested objects typed incorrectly.

Understand the failure class

Background: Schema validation failed / invalid input schema: payload rejected because its shape doesn't match the expected schema — this error's family across 28 libraries.

Related errors

• {validation_error_msg} Pydantic validation error: {e}
• The input list is empty
• The dict format is invalid for this route type. Ensure the s
• One or more lists in the returned prediction response are em
• Invalid data for {dataclass_class.__name__}: {e}

AI-assisted analysis of mlflow/mlflow@6a27f2decc (2026-08-29). Data as JSON: /api/errors/f30ad60b08322294. Report an issue: GitHub.