Fix pydantic root_validator usage

[thinkall]

Why are these changes needed?

pydantic new version is released today, thus current code will raise below error:

https://github.com/microsoft/FLAML/actions/runs/5428592991/jobs/9872924665?pr=1097#step:12:98

ERROR test/autogen/test_math_user_proxy_agent.py - pydantic.errors.PydanticUserError: If you use `@root_validator` with pre=False (the default) you MUST specify `skip_on_failure=True`. Note that `@root_validator` is deprecated and should be replaced with `@model_validator`.

This PR added skip_on_failure=True to fix the error.

Related issue number

Checks

• I've included any doc changes needed for https://microsoft.github.io/FLAML/. See https://microsoft.github.io/FLAML/docs/Contribute#documentation to build and test documentation locally.
• I've added tests (if relevant) corresponding to the changes introduced in this PR.
• I've made sure all auto checks have passed.

[thinkall]

ray has an issue with latest pydantic as well:
https://github.com/microsoft/FLAML/actions/runs/5428749030/jobs/9873215390?pr=1105#step:12:3174

_______________________________ test_consistency _______________________________

self = <ray.tune.execution.trial_runner.TrialRunner object at 0x7f02a18d9970>
next_trial = evaluate_config_a4cb7aa6

    def _wait_and_handle_event(self, next_trial: Optional[Trial]):
        try:
            # Single wait of entire tune loop.
>           event = self.trial_executor.get_next_executor_event(
                self._live_trials, next_trial is not None
            )

/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/tune/execution/trial_runner.py:1369: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/tune/execution/ray_trial_executor.py:1119: in get_next_executor_event
    self._stage_and_update_status(live_trials)
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/tune/execution/ray_trial_executor.py:318: in _stage_and_update_status
    self._resource_manager.request_resources(resource_request=resource_request)
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/air/execution/resources/placement_group.py:143: in request_resources
    future = pg.ready()
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/util/placement_group.py:78: in ready
    return bundle_reservation_check.options(
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/remote_function.py:225: in remote
    return func_cls._remote(args=args, kwargs=kwargs, **updated_options)
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/util/tracing/tracing_helper.py:301: in _invocation_remote_span
    return method(self, args, kwargs, *_args, **_kwargs)
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/remote_function.py:405: in _remote
    return invocation(args, kwargs)
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/remote_function.py:380: in invocation
    object_refs = worker.core_worker.submit_task(
python/ray/_raylet.pyx:2008: in ray._raylet.CoreWorker.submit_task
    ???
python/ray/_raylet.pyx:2012: in ray._raylet.CoreWorker.submit_task
    ???
python/ray/_raylet.pyx:427: in ray._raylet.prepare_args_and_increment_put_refs
    ???
python/ray/_raylet.pyx:418: in ray._raylet.prepare_args_and_increment_put_refs
    ???
python/ray/_raylet.pyx:464: in ray._raylet.prepare_args_internal
    ???
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/_private/worker.py:601: in get_serialization_context
    context_map[job_id] = serialization.SerializationContext(self)
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/_private/serialization.py:151: in __init__
    serialization_addons.apply(self)
/opt/hostedtoolcache/Python/3.9.17/x64/lib/python3.9/site-packages/ray/util/serialization_addons.py:58: in apply
    register_pydantic_serializer(serialization_context)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

serialization_context = <ray._private.serialization.SerializationContext object at 0x7f02a18da430>

    @DeveloperAPI
    def register_pydantic_serializer(serialization_context):
        try:
            import pydantic.fields
        except ImportError:
            return
    
        # Pydantic's Cython validators are not serializable.
        # https://github.com/cloudpipe/cloudpickle/issues/408
        serialization_context._register_cloudpickle_serializer(
>           pydantic.fields.ModelField,
            custom_serializer=lambda o: {
                "name": o.name,
                # outer_type_ is the original type for ModelFields,
                # while type_ can be updated later with the nested type
                # like int for List[int].
                "type_": o.outer_type_,
                "class_validators": o.class_validators,
                "model_config": o.model_config,
                "default": o.default,
                "default_factory": o.default_factory,
                "required": o.required,
                "alias": o.alias,
                "field_info": o.field_info,
            },
            custom_deserializer=lambda kwargs: pydantic.fields.ModelField(**kwargs),
        )
E       AttributeError: module 'pydantic.fields' has no attribute 'ModelField'

[yiranwu0]

Yes, pydantic just updated to the new version of v2, I think we can stay with v1 for now

[thinkall]

Yes, pydantic just updated to the new version of v2, I think we can stay with v1 for now

I see you already fixed the pydantic version in another PR. So, I'll close this one.