[Question]: SchemaLLMPathExtractor w/ Sonnet Fails for some Schemas

[kevin-yang-racap]

Question Validation

• I have searched both the documentation and discord for an answer.

Question

Hi, I am trying to get a SchemaLLMPathExtractor set up but I keep running into an assertion error with the kg_schema. The issue was actually the kg_schema is set to ```
'1 validation error for KGSchema
triplets
Field required [type=missing, input_value={}, input_type=dict]
For further information visit https://errors.pydantic.dev/2.9/v/missing'

is there anything I can do to fix this error?

I am using AWS Bedrock Claude 3 Sonnet.

I tried with a simpler schema and it works fine.

[dosubot]

To fix the assertion error related to the missing 'triplets' field in the kg_schema, ensure that the kg_schema_cls is correctly initialized with the triplets field. The triplets field should be a list of triplet_cls instances. Here's a snippet of the relevant code setup:

triplet_cls = create_model(
    "Triplet",
    subject=(entity_cls, ...),
    relation=(relation_cls, ...),
    object=(entity_cls, ...),
)

root = field_validator("triplets", mode="before")(validate)
kg_schema_cls = create_model(
    "KGSchema",
    __validators__={"validator1": root},  # type: ignore
    triplets=(List[triplet_cls], ...),  # type: ignore
)

Make sure that when you instantiate kg_schema_cls, you provide a list of triplets that conform to the triplet_cls model. This should resolve the assertion error by ensuring the triplets field is present and correctly formatted ^[1].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[logan-markewich]

This quite literally means the llm did not output the proper/expected schema

Maybe try a different extractor

[kevin-yang-racap]

I want to use the SchemaLLMPathExtractor's fixed schema rather than use the DynamicLLMPathExtractor.

Does the pydantic check for partially filled out schema? Is there a way to see what the LLM returned to debug?