vellum-ai 0.8.12__tar.gz → 0.8.14__tar.gz

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vellum-ai
-Version: 0.8.12
+Version: 0.8.14
 Summary: 
 License: MIT
 Requires-Python: >=3.8,<4.0

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "vellum-ai"
-version = "0.8.12"
+version = "0.8.14"
 description = ""
 readme = "README.md"
 authors = []

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/src/vellum/client.py

@@ -19,6 +19,7 @@ from .types.prompt_deployment_expand_meta_request import PromptDeploymentExpandM
 from .types.raw_prompt_execution_overrides_request import RawPromptExecutionOverridesRequest
 from .core.request_options import RequestOptions
 from .types.execute_prompt_response import ExecutePromptResponse
+from .core.serialization import convert_and_respect_annotation_metadata
 from .core.pydantic_utilities import parse_obj_as
 from .errors.bad_request_error import BadRequestError
 from .errors.forbidden_error import ForbiddenError
@@ -197,13 +198,19 @@ class Vellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "inputs": inputs,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[PromptDeploymentInputRequest], direction="write"
+                ),
                 "prompt_deployment_id": prompt_deployment_id,
                 "prompt_deployment_name": prompt_deployment_name,
                 "release_tag": release_tag,
                 "external_id": external_id,
-                "expand_meta": expand_meta,
-                "raw_overrides": raw_overrides,
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=PromptDeploymentExpandMetaRequest, direction="write"
+                ),
+                "raw_overrides": convert_and_respect_annotation_metadata(
+                    object_=raw_overrides, annotation=RawPromptExecutionOverridesRequest, direction="write"
+                ),
                 "expand_raw": expand_raw,
                 "metadata": metadata,
             },
@@ -366,13 +373,19 @@ class Vellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "inputs": inputs,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[PromptDeploymentInputRequest], direction="write"
+                ),
                 "prompt_deployment_id": prompt_deployment_id,
                 "prompt_deployment_name": prompt_deployment_name,
                 "release_tag": release_tag,
                 "external_id": external_id,
-                "expand_meta": expand_meta,
-                "raw_overrides": raw_overrides,
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=PromptDeploymentExpandMetaRequest, direction="write"
+                ),
+                "raw_overrides": convert_and_respect_annotation_metadata(
+                    object_=raw_overrides, annotation=RawPromptExecutionOverridesRequest, direction="write"
+                ),
                 "expand_raw": expand_raw,
                 "metadata": metadata,
             },
@@ -508,8 +521,12 @@ class Vellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "inputs": inputs,
-                "expand_meta": expand_meta,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[WorkflowRequestInputRequest], direction="write"
+                ),
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=WorkflowExpandMetaRequest, direction="write"
+                ),
                 "workflow_deployment_id": workflow_deployment_id,
                 "workflow_deployment_name": workflow_deployment_name,
                 "release_tag": release_tag,
@@ -649,8 +666,12 @@ class Vellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "inputs": inputs,
-                "expand_meta": expand_meta,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[WorkflowRequestInputRequest], direction="write"
+                ),
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=WorkflowExpandMetaRequest, direction="write"
+                ),
                 "workflow_deployment_id": workflow_deployment_id,
                 "workflow_deployment_name": workflow_deployment_name,
                 "release_tag": release_tag,
@@ -772,8 +793,12 @@ class Vellum:
             json={
                 "deployment_id": deployment_id,
                 "deployment_name": deployment_name,
-                "requests": requests,
-                "options": options,
+                "requests": convert_and_respect_annotation_metadata(
+                    object_=requests, annotation=typing.Sequence[GenerateRequest], direction="write"
+                ),
+                "options": convert_and_respect_annotation_metadata(
+                    object_=options, annotation=GenerateOptionsRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -915,8 +940,12 @@ class Vellum:
             json={
                 "deployment_id": deployment_id,
                 "deployment_name": deployment_name,
-                "requests": requests,
-                "options": options,
+                "requests": convert_and_respect_annotation_metadata(
+                    object_=requests, annotation=typing.Sequence[GenerateRequest], direction="write"
+                ),
+                "options": convert_and_respect_annotation_metadata(
+                    object_=options, annotation=GenerateOptionsRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -1036,7 +1065,9 @@ class Vellum:
                 "index_id": index_id,
                 "index_name": index_name,
                 "query": query,
-                "options": options,
+                "options": convert_and_respect_annotation_metadata(
+                    object_=options, annotation=SearchRequestOptionsRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -1132,7 +1163,9 @@ class Vellum:
             json={
                 "deployment_id": deployment_id,
                 "deployment_name": deployment_name,
-                "actuals": actuals,
+                "actuals": convert_and_respect_annotation_metadata(
+                    object_=actuals, annotation=typing.Sequence[SubmitCompletionActualRequest], direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -1222,7 +1255,9 @@ class Vellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "actuals": actuals,
+                "actuals": convert_and_respect_annotation_metadata(
+                    object_=actuals, annotation=typing.Sequence[SubmitWorkflowExecutionActualRequest], direction="write"
+                ),
                 "execution_id": execution_id,
                 "external_id": external_id,
             },
@@ -1386,13 +1421,19 @@ class AsyncVellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "inputs": inputs,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[PromptDeploymentInputRequest], direction="write"
+                ),
                 "prompt_deployment_id": prompt_deployment_id,
                 "prompt_deployment_name": prompt_deployment_name,
                 "release_tag": release_tag,
                 "external_id": external_id,
-                "expand_meta": expand_meta,
-                "raw_overrides": raw_overrides,
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=PromptDeploymentExpandMetaRequest, direction="write"
+                ),
+                "raw_overrides": convert_and_respect_annotation_metadata(
+                    object_=raw_overrides, annotation=RawPromptExecutionOverridesRequest, direction="write"
+                ),
                 "expand_raw": expand_raw,
                 "metadata": metadata,
             },
@@ -1563,13 +1604,19 @@ class AsyncVellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "inputs": inputs,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[PromptDeploymentInputRequest], direction="write"
+                ),
                 "prompt_deployment_id": prompt_deployment_id,
                 "prompt_deployment_name": prompt_deployment_name,
                 "release_tag": release_tag,
                 "external_id": external_id,
-                "expand_meta": expand_meta,
-                "raw_overrides": raw_overrides,
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=PromptDeploymentExpandMetaRequest, direction="write"
+                ),
+                "raw_overrides": convert_and_respect_annotation_metadata(
+                    object_=raw_overrides, annotation=RawPromptExecutionOverridesRequest, direction="write"
+                ),
                 "expand_raw": expand_raw,
                 "metadata": metadata,
             },
@@ -1713,8 +1760,12 @@ class AsyncVellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "inputs": inputs,
-                "expand_meta": expand_meta,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[WorkflowRequestInputRequest], direction="write"
+                ),
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=WorkflowExpandMetaRequest, direction="write"
+                ),
                 "workflow_deployment_id": workflow_deployment_id,
                 "workflow_deployment_name": workflow_deployment_name,
                 "release_tag": release_tag,
@@ -1862,8 +1913,12 @@ class AsyncVellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "inputs": inputs,
-                "expand_meta": expand_meta,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[WorkflowRequestInputRequest], direction="write"
+                ),
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=WorkflowExpandMetaRequest, direction="write"
+                ),
                 "workflow_deployment_id": workflow_deployment_id,
                 "workflow_deployment_name": workflow_deployment_name,
                 "release_tag": release_tag,
@@ -1993,8 +2048,12 @@ class AsyncVellum:
             json={
                 "deployment_id": deployment_id,
                 "deployment_name": deployment_name,
-                "requests": requests,
-                "options": options,
+                "requests": convert_and_respect_annotation_metadata(
+                    object_=requests, annotation=typing.Sequence[GenerateRequest], direction="write"
+                ),
+                "options": convert_and_respect_annotation_metadata(
+                    object_=options, annotation=GenerateOptionsRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -2144,8 +2203,12 @@ class AsyncVellum:
             json={
                 "deployment_id": deployment_id,
                 "deployment_name": deployment_name,
-                "requests": requests,
-                "options": options,
+                "requests": convert_and_respect_annotation_metadata(
+                    object_=requests, annotation=typing.Sequence[GenerateRequest], direction="write"
+                ),
+                "options": convert_and_respect_annotation_metadata(
+                    object_=options, annotation=GenerateOptionsRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -2273,7 +2336,9 @@ class AsyncVellum:
                 "index_id": index_id,
                 "index_name": index_name,
                 "query": query,
-                "options": options,
+                "options": convert_and_respect_annotation_metadata(
+                    object_=options, annotation=SearchRequestOptionsRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -2377,7 +2442,9 @@ class AsyncVellum:
             json={
                 "deployment_id": deployment_id,
                 "deployment_name": deployment_name,
-                "actuals": actuals,
+                "actuals": convert_and_respect_annotation_metadata(
+                    object_=actuals, annotation=typing.Sequence[SubmitCompletionActualRequest], direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -2475,7 +2542,9 @@ class AsyncVellum:
             base_url=self._client_wrapper.get_environment().predict,
             method="POST",
             json={
-                "actuals": actuals,
+                "actuals": convert_and_respect_annotation_metadata(
+                    object_=actuals, annotation=typing.Sequence[SubmitWorkflowExecutionActualRequest], direction="write"
+                ),
                 "execution_id": execution_id,
                 "external_id": external_id,
             },

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/src/vellum/core/__init__.py

@@ -3,7 +3,7 @@
 from .api_error import ApiError
 from .client_wrapper import AsyncClientWrapper, BaseClientWrapper, SyncClientWrapper
 from .datetime_utils import serialize_datetime
-from .file import File, convert_file_dict_to_httpx_tuples
+from .file import File, convert_file_dict_to_httpx_tuples, with_content_type
 from .http_client import AsyncHttpClient, HttpClient
 from .jsonable_encoder import jsonable_encoder
 from .pydantic_utilities import (
@@ -43,4 +43,5 @@ __all__ = [
     "universal_field_validator",
     "universal_root_validator",
     "update_forward_refs",
+    "with_content_type",
 ]

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/src/vellum/core/client_wrapper.py

@@ -17,7 +17,7 @@ class BaseClientWrapper:
         headers: typing.Dict[str, str] = {
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "vellum-ai",
-            "X-Fern-SDK-Version": "0.8.12",
+            "X-Fern-SDK-Version": "0.8.14",
         }
         headers["X_API_KEY"] = self.api_key
         return headers

vellum_ai-0.8.14/src/vellum/core/file.py

@@ -0,0 +1,62 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from typing import IO, Dict, List, Mapping, Optional, Tuple, Union, cast
+
+# File typing inspired by the flexibility of types within the httpx library
+# https://github.com/encode/httpx/blob/master/httpx/_types.py
+FileContent = Union[IO[bytes], bytes, str]
+File = Union[
+    # file (or bytes)
+    FileContent,
+    # (filename, file (or bytes))
+    Tuple[Optional[str], FileContent],
+    # (filename, file (or bytes), content_type)
+    Tuple[Optional[str], FileContent, Optional[str]],
+    # (filename, file (or bytes), content_type, headers)
+    Tuple[
+        Optional[str],
+        FileContent,
+        Optional[str],
+        Mapping[str, str],
+    ],
+]
+
+
+def convert_file_dict_to_httpx_tuples(
+    d: Dict[str, Union[File, List[File]]],
+) -> List[Tuple[str, File]]:
+    """
+    The format we use is a list of tuples, where the first element is the
+    name of the file and the second is the file object. Typically HTTPX wants
+    a dict, but to be able to send lists of files, you have to use the list
+    approach (which also works for non-lists)
+    https://github.com/encode/httpx/pull/1032
+    """
+
+    httpx_tuples = []
+    for key, file_like in d.items():
+        if isinstance(file_like, list):
+            for file_like_item in file_like:
+                httpx_tuples.append((key, file_like_item))
+        else:
+            httpx_tuples.append((key, file_like))
+    return httpx_tuples
+
+
+def with_content_type(*, file: File, content_type: str) -> File:
+    """ """
+    if isinstance(file, tuple):
+        if len(file) == 2:
+            filename, content = cast(Tuple[Optional[str], FileContent], file)  # type: ignore
+            return (filename, content, content_type)
+        elif len(file) == 3:
+            filename, content, _ = cast(Tuple[Optional[str], FileContent, Optional[str]], file)  # type: ignore
+            return (filename, content, content_type)
+        elif len(file) == 4:
+            filename, content, _, headers = cast(  # type: ignore
+                Tuple[Optional[str], FileContent, Optional[str], Mapping[str, str]], file
+            )
+            return (filename, content, content_type, headers)
+        else:
+            raise ValueError(f"Unexpected tuple length: {len(file)}")
+    return (None, file, content_type)

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/src/vellum/core/http_client.py

@@ -224,7 +224,9 @@ class HttpClient:
             json=json_body,
             data=data_body,
             content=content,
-            files=convert_file_dict_to_httpx_tuples(remove_none_from_dict(files)) if files is not None else None,
+            files=convert_file_dict_to_httpx_tuples(remove_none_from_dict(files))
+            if (files is not None and files is not omit)
+            else None,
             timeout=timeout,
         )
 
@@ -306,7 +308,9 @@ class HttpClient:
             json=json_body,
             data=data_body,
             content=content,
-            files=convert_file_dict_to_httpx_tuples(remove_none_from_dict(files)) if files is not None else None,
+            files=convert_file_dict_to_httpx_tuples(remove_none_from_dict(files))
+            if (files is not None and files is not omit)
+            else None,
             timeout=timeout,
         ) as stream:
             yield stream

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/src/vellum/core/pydantic_utilities.py

@@ -10,6 +10,7 @@ import typing_extensions
 import pydantic
 
 from .datetime_utils import serialize_datetime
+from .serialization import convert_and_respect_annotation_metadata
 
 IS_PYDANTIC_V2 = pydantic.VERSION.startswith("2.")
 
@@ -56,11 +57,12 @@ Model = typing.TypeVar("Model", bound=pydantic.BaseModel)
 
 
 def parse_obj_as(type_: typing.Type[T], object_: typing.Any) -> T:
+    dealiased_object = convert_and_respect_annotation_metadata(object_=object_, annotation=type_, direction="read")
     if IS_PYDANTIC_V2:
         adapter = pydantic.TypeAdapter(type_)  # type: ignore # Pydantic v2
-        return adapter.validate_python(object_)
+        return adapter.validate_python(dealiased_object)
     else:
-        return pydantic.parse_obj_as(type_, object_)
+        return pydantic.parse_obj_as(type_, dealiased_object)
 
 
 def to_jsonable_with_fallback(
@@ -75,13 +77,40 @@ def to_jsonable_with_fallback(
 
 
 class UniversalBaseModel(pydantic.BaseModel):
-    class Config:
-        populate_by_name = True
-        smart_union = True
-        allow_population_by_field_name = True
-        json_encoders = {dt.datetime: serialize_datetime}
-        # Allow fields begining with `model_` to be used in the model
-        protected_namespaces = ()
+    if IS_PYDANTIC_V2:
+        model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(
+            # Allow fields begining with `model_` to be used in the model
+            protected_namespaces=(),
+        )  # type: ignore # Pydantic v2
+
+        @pydantic.model_serializer(mode="wrap", when_used="json")  # type: ignore # Pydantic v2
+        def serialize_model(self, handler: pydantic.SerializerFunctionWrapHandler) -> typing.Any:  # type: ignore # Pydantic v2
+            serialized = handler(self)
+            data = {k: serialize_datetime(v) if isinstance(v, dt.datetime) else v for k, v in serialized.items()}
+            return data
+
+    else:
+
+        class Config:
+            smart_union = True
+            json_encoders = {dt.datetime: serialize_datetime}
+
+    @classmethod
+    def model_construct(
+        cls: typing.Type["Model"], _fields_set: typing.Optional[typing.Set[str]] = None, **values: typing.Any
+    ) -> "Model":
+        dealiased_object = convert_and_respect_annotation_metadata(object_=values, annotation=cls, direction="read")
+        return cls.construct(_fields_set, **dealiased_object)
+
+    @classmethod
+    def construct(
+        cls: typing.Type["Model"], _fields_set: typing.Optional[typing.Set[str]] = None, **values: typing.Any
+    ) -> "Model":
+        dealiased_object = convert_and_respect_annotation_metadata(object_=values, annotation=cls, direction="read")
+        if IS_PYDANTIC_V2:
+            return super().model_construct(_fields_set, **dealiased_object)  # type: ignore # Pydantic v2
+        else:
+            return super().construct(_fields_set, **dealiased_object)
 
     def json(self, **kwargs: typing.Any) -> str:
         kwargs_with_defaults: typing.Any = {
@@ -117,7 +146,7 @@ class UniversalBaseModel(pydantic.BaseModel):
                 "exclude_none": True,
                 "exclude_unset": False,
             }
-            return deep_union_pydantic_dicts(
+            dict_dump = deep_union_pydantic_dicts(
                 super().model_dump(**kwargs_with_defaults_exclude_unset),  # type: ignore # Pydantic v2
                 super().model_dump(**kwargs_with_defaults_exclude_none),  # type: ignore # Pydantic v2
             )
@@ -143,7 +172,9 @@ class UniversalBaseModel(pydantic.BaseModel):
                 **kwargs,
             }
 
-            return super().dict(**kwargs_with_defaults_exclude_unset_include_fields)
+            dict_dump = super().dict(**kwargs_with_defaults_exclude_unset_include_fields)
+
+        return convert_and_respect_annotation_metadata(object_=dict_dump, annotation=self.__class__, direction="write")
 
 
 def deep_union_pydantic_dicts(
@@ -183,11 +214,11 @@ def encode_by_type(o: typing.Any) -> typing.Any:
             return encoder(o)
 
 
-def update_forward_refs(model: typing.Type["Model"]) -> None:
+def update_forward_refs(model: typing.Type["Model"], **localns: typing.Any) -> None:
     if IS_PYDANTIC_V2:
         model.model_rebuild(raise_errors=False)  # type: ignore # Pydantic v2
     else:
-        model.update_forward_refs()
+        model.update_forward_refs(**localns)
 
 
 # Mirrors Pydantic's internal typing

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/src/vellum/core/serialization.py

@@ -71,6 +71,24 @@ def convert_and_respect_annotation_metadata(
     if typing_extensions.is_typeddict(clean_type) and isinstance(object_, typing.Mapping):
         return _convert_mapping(object_, clean_type, direction)
 
+    if (
+        typing_extensions.get_origin(clean_type) == typing.Dict
+        or typing_extensions.get_origin(clean_type) == dict
+        or clean_type == typing.Dict
+    ) and isinstance(object_, typing.Dict):
+        key_type = typing_extensions.get_args(clean_type)[0]
+        value_type = typing_extensions.get_args(clean_type)[1]
+
+        return {
+            key: convert_and_respect_annotation_metadata(
+                object_=value,
+                annotation=annotation,
+                inner_type=value_type,
+                direction=direction,
+            )
+            for key, value in object_.items()
+        }
+
     # If you're iterating on a string, do not bother to coerce it to a sequence.
     if not isinstance(object_, str):
         if (

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/src/vellum/resources/ad_hoc/client.py

@@ -9,6 +9,7 @@ from ...types.prompt_block_request import PromptBlockRequest
 from ...types.ad_hoc_expand_meta_request import AdHocExpandMetaRequest
 from ...core.request_options import RequestOptions
 from ...types.ad_hoc_execute_prompt_event import AdHocExecutePromptEvent
+from ...core.serialization import convert_and_respect_annotation_metadata
 from ...core.pydantic_utilities import parse_obj_as
 import json
 from ...errors.bad_request_error import BadRequestError
@@ -133,11 +134,21 @@ class AdHocClient:
             method="POST",
             json={
                 "ml_model": ml_model,
-                "input_values": input_values,
-                "input_variables": input_variables,
-                "parameters": parameters,
-                "blocks": blocks,
-                "expand_meta": expand_meta,
+                "input_values": convert_and_respect_annotation_metadata(
+                    object_=input_values, annotation=typing.Sequence[PromptRequestInputRequest], direction="write"
+                ),
+                "input_variables": convert_and_respect_annotation_metadata(
+                    object_=input_variables, annotation=typing.Sequence[VellumVariableRequest], direction="write"
+                ),
+                "parameters": convert_and_respect_annotation_metadata(
+                    object_=parameters, annotation=PromptParametersRequest, direction="write"
+                ),
+                "blocks": convert_and_respect_annotation_metadata(
+                    object_=blocks, annotation=typing.Sequence[PromptBlockRequest], direction="write"
+                ),
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=AdHocExpandMetaRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -314,11 +325,21 @@ class AsyncAdHocClient:
             method="POST",
             json={
                 "ml_model": ml_model,
-                "input_values": input_values,
-                "input_variables": input_variables,
-                "parameters": parameters,
-                "blocks": blocks,
-                "expand_meta": expand_meta,
+                "input_values": convert_and_respect_annotation_metadata(
+                    object_=input_values, annotation=typing.Sequence[PromptRequestInputRequest], direction="write"
+                ),
+                "input_variables": convert_and_respect_annotation_metadata(
+                    object_=input_variables, annotation=typing.Sequence[VellumVariableRequest], direction="write"
+                ),
+                "parameters": convert_and_respect_annotation_metadata(
+                    object_=parameters, annotation=PromptParametersRequest, direction="write"
+                ),
+                "blocks": convert_and_respect_annotation_metadata(
+                    object_=blocks, annotation=typing.Sequence[PromptBlockRequest], direction="write"
+                ),
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=AdHocExpandMetaRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,

{vellum_ai-0.8.12 → vellum_ai-0.8.14}/src/vellum/resources/deployments/client.py

@@ -14,6 +14,7 @@ from ...types.deployment_release_tag_read import DeploymentReleaseTagRead
 from ...types.prompt_deployment_input_request import PromptDeploymentInputRequest
 from ...types.compile_prompt_deployment_expand_meta_request import CompilePromptDeploymentExpandMetaRequest
 from ...types.deployment_provider_payload_response import DeploymentProviderPayloadResponse
+from ...core.serialization import convert_and_respect_annotation_metadata
 from ...errors.bad_request_error import BadRequestError
 from ...errors.forbidden_error import ForbiddenError
 from ...errors.not_found_error import NotFoundError
@@ -334,9 +335,13 @@ class DeploymentsClient:
             json={
                 "deployment_id": deployment_id,
                 "deployment_name": deployment_name,
-                "inputs": inputs,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[PromptDeploymentInputRequest], direction="write"
+                ),
                 "release_tag": release_tag,
-                "expand_meta": expand_meta,
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=CompilePromptDeploymentExpandMetaRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,
@@ -746,9 +751,13 @@ class AsyncDeploymentsClient:
             json={
                 "deployment_id": deployment_id,
                 "deployment_name": deployment_name,
-                "inputs": inputs,
+                "inputs": convert_and_respect_annotation_metadata(
+                    object_=inputs, annotation=typing.Sequence[PromptDeploymentInputRequest], direction="write"
+                ),
                 "release_tag": release_tag,
-                "expand_meta": expand_meta,
+                "expand_meta": convert_and_respect_annotation_metadata(
+                    object_=expand_meta, annotation=CompilePromptDeploymentExpandMetaRequest, direction="write"
+                ),
             },
             request_options=request_options,
             omit=OMIT,