Pydantic validation error when

At a glance

The community member is experiencing a Pydantic validation error when attempting to subclass ChatResponse from llama_index.core.llms and extend it with a custom Pydantic model. The error is thrown when they add a list of MyModels to ChatResponse via subclassing. However, the error is not thrown if they don't subclass from ChatResponse.

The community member can resolve the error by importing BaseModel from llama_index.core.bridge.pydantic, but they find this approach to be very clunky as they are importing BaseModel from Pydantic elsewhere in their application.

The comments suggest that the issue may be due to the handling of Pydantic versions within the llama-index library. Specifically, version 0.10.x of llama-index is using Pydantic v1, so the community member's subclass should also import from there (from pydantic.v1 import BaseModel). The comments also mention that version 0.11.x of

aalfredmadere

Pydantic validation error when attempting to subclass ChatResponse from llama_index.core.llms and extend it with a custom pydantic model.

The error is thrown when I add a list of MyModels to ChatResponse via subclassing like so

Plain Text

from pydantic import BaseModel
from llama_index.core.llms import ChatResponse
class MyModel(BaseModel):
  foo: str

class CompletionWithAttributions(ChatResponse):
  attributions: list[MyModel] | None = None

The error is not thrown if I don't subclass from ChatResponse like so

Plain Text

from pydantic import BaseModel
class MyModel(BaseModel):
  foo: str

class CompletionWithAttributions(BaseModel):
  attributions: list[MyModel] | None = None

I can resolve this error by importing BaseModel from llama-index itself like this, but that seems VERY clugy as elsewhere in my app i am importing BaseModel from pydantic as per usual.

Plain Text

from llama_index.core.bridge.pydantic import BaseModel

Anyone experienced anything like this with llama-index? I suspect its due to the handling of pydantic versions within the library.

I am using
llama-index==0.10.51
pydantic==2.6.4

3 comments

LLogan M

0.10.x of llama-index is using pydantic.v1

LLogan M

so your subclass should also import from there

from pydantic.v1 import BaseModel

LLogan M

v0.11.x updates to pydantic v2 fully though, heads up

Add a reply

Find answers from the community

Pydantic validation error when