mistralai 1.0.0rc1__py3-none-any.whl → 1.0.1__py3-none-any.whl

mistralai_azure/sdk.py

@@ -3,8 +3,7 @@
 from typing import Any, Callable, Dict, Optional, Union
 
 import httpx
-import mistralai_azure.utils as utils
-from mistralai_azure import models
+from mistralai_azure import models, utils
 from mistralai_azure._hooks import SDKHooks
 from mistralai_azure.chat import Chat
 from mistralai_azure.types import Nullable
@@ -12,6 +11,7 @@ from mistralai_azure.types import Nullable
 from .basesdk import BaseSDK
 from .httpclient import AsyncHttpClient, HttpClient
 from .sdkconfiguration import SDKConfiguration
+from .utils.logger import Logger, NoOpLogger
 from .utils.retries import RetryConfig
 
 
@@ -29,6 +29,7 @@ class MistralAzure(BaseSDK):
         client: Optional[HttpClient] = None,
         async_client: Optional[AsyncHttpClient] = None,
         retry_config: Optional[Nullable[RetryConfig]] = None,
+        debug_logger: Optional[Logger] = None,
     ) -> None:
         r"""Instantiates the SDK configuring it with the provided parameters.
 
@@ -60,6 +61,9 @@ class MistralAzure(BaseSDK):
             type(async_client), AsyncHttpClient
         ), "The provided async_client must implement the AsyncHttpClient protocol."
 
+        if debug_logger is None:
+            debug_logger = NoOpLogger()
+
         security: Any = None
         if callable(azure_api_key):
             security = lambda: models.Security(  # pylint: disable=unnecessary-lambda-assignment
@@ -81,6 +85,7 @@ class MistralAzure(BaseSDK):
                 server_url=server_url,
                 server=None,
                 retry_config=retry_config,
+                debug_logger=debug_logger,
             ),
         )
 

mistralai_azure/sdkconfiguration.py

@@ -3,7 +3,7 @@
 
 from ._hooks import SDKHooks
 from .httpclient import AsyncHttpClient, HttpClient
-from .utils import RetryConfig, remove_suffix
+from .utils import Logger, RetryConfig, remove_suffix
 from dataclasses import dataclass
 from mistralai_azure import models
 from mistralai_azure.types import OptionalNullable, UNSET
@@ -23,14 +23,15 @@ SERVERS = {
 class SDKConfiguration:
     client: HttpClient
     async_client: AsyncHttpClient
+    debug_logger: Logger
     security: Optional[Union[models.Security,Callable[[], models.Security]]] = None
     server_url: Optional[str] = ""
     server: Optional[str] = ""
     language: str = "python"
     openapi_doc_version: str = "0.0.2"
-    sdk_version: str = "1.0.0rc1"
-    gen_version: str = "2.382.2"
-    user_agent: str = "speakeasy-sdk/python 1.0.0rc1 2.382.2 0.0.2 mistralai_azure"
+    sdk_version: str = "1.0.0-rc.4"
+    gen_version: str = "2.390.6"
+    user_agent: str = "speakeasy-sdk/python 1.0.0-rc.4 2.390.6 0.0.2 mistralai_azure"
     retry_config: OptionalNullable[RetryConfig] = Field(default_factory=lambda: UNSET)
     timeout_ms: Optional[int] = None
 

mistralai_azure/types/basemodel.py

@@ -2,8 +2,8 @@
 
 from pydantic import ConfigDict, model_serializer
 from pydantic import BaseModel as PydanticBaseModel
-from typing import Literal, Optional, TypeVar, Union, NewType
-from typing_extensions import TypeAliasType
+from typing import TYPE_CHECKING, Literal, Optional, TypeVar, Union, NewType
+from typing_extensions import TypeAliasType, TypeAlias
 
 
 class BaseModel(PydanticBaseModel):
@@ -26,10 +26,14 @@ UNSET_SENTINEL = "~?~unset~?~sentinel~?~"
 
 
 T = TypeVar("T")
-Nullable = TypeAliasType("Nullable", Union[T, None], type_params=(T,))
-OptionalNullable = TypeAliasType(
-    "OptionalNullable", Union[Optional[Nullable[T]], Unset], type_params=(T,)
-)
+if TYPE_CHECKING:
+    Nullable: TypeAlias = Union[T, None]
+    OptionalNullable: TypeAlias = Union[Optional[Nullable[T]], Unset]
+else:
+    Nullable = TypeAliasType("Nullable", Union[T, None], type_params=(T,))
+    OptionalNullable = TypeAliasType(
+        "OptionalNullable", Union[Optional[Nullable[T]], Unset], type_params=(T,)
+    )
 
 UnrecognizedInt = NewType("UnrecognizedInt", int)
 UnrecognizedStr = NewType("UnrecognizedStr", str)

mistralai_azure/utils/__init__.py

@@ -34,6 +34,7 @@ from .serializers import (
 )
 from .url import generate_url, template_url, remove_suffix
 from .values import get_global_from_env, match_content_type, match_status_codes, match_response
+from .logger import Logger, get_body_content, NoOpLogger
 
 __all__ = [
     "BackoffStrategy",
@@ -41,6 +42,7 @@ __all__ = [
     "find_metadata",
     "FormMetadata",
     "generate_url",
+    "get_body_content",
     "get_discriminator",
     "get_global_from_env",
     "get_headers",
@@ -49,11 +51,13 @@ __all__ = [
     "get_response_headers",
     "get_security",
     "HeaderMetadata",
+    "Logger",
     "marshal_json",
     "match_content_type",
     "match_status_codes",
     "match_response",
     "MultipartFormMetadata",
+    "NoOpLogger",
     "OpenEnumMeta",
     "PathParamMetadata",
     "QueryParamMetadata",

mistralai_azure/utils/eventstreaming.py

@@ -147,15 +147,14 @@ def _parse_event(
         data = data[:-1]
         event.data = data
 
-        if (
-            data.isnumeric()
-            or data == "true"
-            or data == "false"
-            or data == "null"
-            or data.startswith("{")
-            or data.startswith("[")
-            or data.startswith('"')
-        ):
+        data_is_primitive = (
+            data.isnumeric() or data == "true" or data == "false" or data == "null"
+        )
+        data_is_json = (
+            data.startswith("{") or data.startswith("[") or data.startswith('"')
+        )
+
+        if data_is_primitive or data_is_json:
             try:
                 event.data = json.loads(data)
             except Exception:

mistralai_azure/utils/logger.py

@@ -0,0 +1,16 @@
+"""Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
+
+import httpx
+from typing import Any, Protocol
+
+class Logger(Protocol):
+    def debug(self, msg: str, *args: Any, **kwargs: Any) -> None:
+        pass
+
+class NoOpLogger:
+    def debug(self, msg: str, *args: Any, **kwargs: Any) -> None:
+        pass
+
+def get_body_content(req: httpx.Request) -> str:
+    return "<streaming body>" if not hasattr(req, "_content") else str(req.content)
+

mistralai_azure/utils/retries.py

@@ -76,7 +76,7 @@ def retry(func, retries: Retries):
 
                         status_major = res.status_code / 100
 
-                        if status_major >= code_range and status_major < code_range + 1:
+                        if code_range <= status_major < code_range + 1:
                             raise TemporaryError(res)
                     else:
                         parsed_code = int(code)
@@ -125,7 +125,7 @@ async def retry_async(func, retries: Retries):
 
                         status_major = res.status_code / 100
 
-                        if status_major >= code_range and status_major < code_range + 1:
+                        if code_range <= status_major < code_range + 1:
                             raise TemporaryError(res)
                     else:
                         parsed_code = int(code)

mistralai_gcp/basesdk.py

@@ -2,10 +2,9 @@
 
 from .sdkconfiguration import SDKConfiguration
 import httpx
-from mistralai_gcp import models
+from mistralai_gcp import models, utils
 from mistralai_gcp._hooks import AfterErrorContext, AfterSuccessContext, BeforeRequestContext
-import mistralai_gcp.utils as utils
-from mistralai_gcp.utils import RetryConfig, SerializedRequestBody
+from mistralai_gcp.utils import RetryConfig, SerializedRequestBody, get_body_content
 from typing import Callable, List, Optional, Tuple
 
 class BaseSDK:
@@ -117,6 +116,7 @@ class BaseSDK:
         retry_config: Optional[Tuple[RetryConfig, List[str]]] = None,
     ) -> httpx.Response:
         client = self.sdk_configuration.client
+        logger = self.sdk_configuration.debug_logger
 
         def do():
             http_res = None
@@ -124,26 +124,45 @@ class BaseSDK:
                 req = self.sdk_configuration.get_hooks().before_request(
                     BeforeRequestContext(hook_ctx), request
                 )
+                logger.debug(
+                    "Request:\nMethod: %s\nURL: %s\nHeaders: %s\nBody: %s",
+                    req.method,
+                    req.url,
+                    req.headers,
+                    get_body_content(req)
+                )
                 http_res = client.send(req, stream=stream)
             except Exception as e:
                 _, e = self.sdk_configuration.get_hooks().after_error(
                     AfterErrorContext(hook_ctx), None, e
                 )
                 if e is not None:
+                    logger.debug("Request Exception", exc_info=True)
                     raise e
 
             if http_res is None:
+                logger.debug("Raising no response SDK error")
                 raise models.SDKError("No response received")
 
+            logger.debug(
+                "Response:\nStatus Code: %s\nURL: %s\nHeaders: %s\nBody: %s",
+                http_res.status_code,
+                http_res.url,
+                http_res.headers,
+                "<streaming response>" if stream else http_res.text
+            )
+
             if utils.match_status_codes(error_status_codes, http_res.status_code):
                 result, err = self.sdk_configuration.get_hooks().after_error(
                     AfterErrorContext(hook_ctx), http_res, None
                 )
                 if err is not None:
+                    logger.debug("Request Exception", exc_info=True)
                     raise err
                 if result is not None:
                     http_res = result
                 else:
+                    logger.debug("Raising unexpected SDK error")
                     raise models.SDKError("Unexpected error occurred")
 
             return http_res
@@ -169,33 +188,52 @@ class BaseSDK:
         retry_config: Optional[Tuple[RetryConfig, List[str]]] = None,
     ) -> httpx.Response:
         client = self.sdk_configuration.async_client
-
+        logger = self.sdk_configuration.debug_logger
         async def do():
             http_res = None
             try:
                 req = self.sdk_configuration.get_hooks().before_request(
                     BeforeRequestContext(hook_ctx), request
                 )
+                logger.debug(
+                    "Request:\nMethod: %s\nURL: %s\nHeaders: %s\nBody: %s",
+                    req.method,
+                    req.url,
+                    req.headers,
+                    get_body_content(req)
+                )
                 http_res = await client.send(req, stream=stream)
             except Exception as e:
                 _, e = self.sdk_configuration.get_hooks().after_error(
                     AfterErrorContext(hook_ctx), None, e
                 )
                 if e is not None:
+                    logger.debug("Request Exception", exc_info=True)
                     raise e
 
             if http_res is None:
+                logger.debug("Raising no response SDK error")
                 raise models.SDKError("No response received")
 
+            logger.debug(
+                "Response:\nStatus Code: %s\nURL: %s\nHeaders: %s\nBody: %s",
+                http_res.status_code,
+                http_res.url,
+                http_res.headers,
+                "<streaming response>" if stream else http_res.text
+            )
+
             if utils.match_status_codes(error_status_codes, http_res.status_code):
                 result, err = self.sdk_configuration.get_hooks().after_error(
                     AfterErrorContext(hook_ctx), http_res, None
                 )
                 if err is not None:
+                    logger.debug("Request Exception", exc_info=True)
                     raise err
                 if result is not None:
                     http_res = result
                 else:
+                    logger.debug("Raising unexpected SDK error")
                     raise models.SDKError("Unexpected error occurred")
 
             return http_res

mistralai_gcp/chat.py

@@ -1,10 +1,9 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from .basesdk import BaseSDK
-from mistralai_gcp import models
+from mistralai_gcp import models, utils
 from mistralai_gcp._hooks import HookContext
-from mistralai_gcp.types import OptionalNullable, UNSET
-import mistralai_gcp.utils as utils
+from mistralai_gcp.types import Nullable, OptionalNullable, UNSET
 from mistralai_gcp.utils import eventstreaming
 from typing import Any, AsyncGenerator, Generator, List, Optional, Union
 
@@ -14,8 +13,8 @@ class Chat(BaseSDK):
     
     def stream(
         self, *,
+        model: Nullable[str],
         messages: Union[List[models.Messages], List[models.MessagesTypedDict]],
-        model: OptionalNullable[str] = UNSET,
         temperature: Optional[float] = 0.7,
         top_p: Optional[float] = 1,
         max_tokens: OptionalNullable[int] = UNSET,
@@ -34,8 +33,8 @@ class Chat(BaseSDK):
 
         Mistral AI provides the ability to stream responses back to a client in order to allow partial results for certain requests. Tokens will be sent as data-only server-sent events as they become available, with the stream terminated by a data: [DONE] message. Otherwise, the server will hold the request open until the timeout or until completion, with the response containing the full result as JSON.
 
-        :param messages: The prompt(s) to generate completions for, encoded as a list of dict with role and content.
         :param model: ID of the model to use. You can use the [List Available Models](/api#operation/listModels) API to see all of your available models, or see our [Model overview](/models) for model descriptions.
+        :param messages: The prompt(s) to generate completions for, encoded as a list of dict with role and content.
         :param temperature: What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both.
         :param top_p: Nucleus sampling, where the model considers the results of the tokens with `top_p` probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.
         :param max_tokens: The maximum number of tokens to generate in the completion. The token count of your prompt plus `max_tokens` cannot exceed the model's context length.
@@ -127,8 +126,8 @@ class Chat(BaseSDK):
     
     async def stream_async(
         self, *,
+        model: Nullable[str],
         messages: Union[List[models.Messages], List[models.MessagesTypedDict]],
-        model: OptionalNullable[str] = UNSET,
         temperature: Optional[float] = 0.7,
         top_p: Optional[float] = 1,
         max_tokens: OptionalNullable[int] = UNSET,
@@ -147,8 +146,8 @@ class Chat(BaseSDK):
 
         Mistral AI provides the ability to stream responses back to a client in order to allow partial results for certain requests. Tokens will be sent as data-only server-sent events as they become available, with the stream terminated by a data: [DONE] message. Otherwise, the server will hold the request open until the timeout or until completion, with the response containing the full result as JSON.
 
-        :param messages: The prompt(s) to generate completions for, encoded as a list of dict with role and content.
         :param model: ID of the model to use. You can use the [List Available Models](/api#operation/listModels) API to see all of your available models, or see our [Model overview](/models) for model descriptions.
+        :param messages: The prompt(s) to generate completions for, encoded as a list of dict with role and content.
         :param temperature: What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both.
         :param top_p: Nucleus sampling, where the model considers the results of the tokens with `top_p` probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.
         :param max_tokens: The maximum number of tokens to generate in the completion. The token count of your prompt plus `max_tokens` cannot exceed the model's context length.
@@ -238,10 +237,10 @@ class Chat(BaseSDK):
 
     
     
-    def create(
+    def complete(
         self, *,
+        model: Nullable[str],
         messages: Union[List[models.ChatCompletionRequestMessages], List[models.ChatCompletionRequestMessagesTypedDict]],
-        model: OptionalNullable[str] = UNSET,
         temperature: Optional[float] = 0.7,
         top_p: Optional[float] = 1,
         max_tokens: OptionalNullable[int] = UNSET,
@@ -258,10 +257,8 @@ class Chat(BaseSDK):
     ) -> Optional[models.ChatCompletionResponse]:
         r"""Chat Completion
 
-        Chat Completion
-
-        :param messages: The prompt(s) to generate completions for, encoded as a list of dict with role and content.
         :param model: ID of the model to use. You can use the [List Available Models](/api#operation/listModels) API to see all of your available models, or see our [Model overview](/models) for model descriptions.
+        :param messages: The prompt(s) to generate completions for, encoded as a list of dict with role and content.
         :param temperature: What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both.
         :param top_p: Nucleus sampling, where the model considers the results of the tokens with `top_p` probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.
         :param max_tokens: The maximum number of tokens to generate in the completion. The token count of your prompt plus `max_tokens` cannot exceed the model's context length.
@@ -350,10 +347,10 @@ class Chat(BaseSDK):
 
     
     
-    async def create_async(
+    async def complete_async(
         self, *,
+        model: Nullable[str],
         messages: Union[List[models.ChatCompletionRequestMessages], List[models.ChatCompletionRequestMessagesTypedDict]],
-        model: OptionalNullable[str] = UNSET,
         temperature: Optional[float] = 0.7,
         top_p: Optional[float] = 1,
         max_tokens: OptionalNullable[int] = UNSET,
@@ -370,10 +367,8 @@ class Chat(BaseSDK):
     ) -> Optional[models.ChatCompletionResponse]:
         r"""Chat Completion
 
-        Chat Completion
-
-        :param messages: The prompt(s) to generate completions for, encoded as a list of dict with role and content.
         :param model: ID of the model to use. You can use the [List Available Models](/api#operation/listModels) API to see all of your available models, or see our [Model overview](/models) for model descriptions.
+        :param messages: The prompt(s) to generate completions for, encoded as a list of dict with role and content.
         :param temperature: What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both.
         :param top_p: Nucleus sampling, where the model considers the results of the tokens with `top_p` probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.
         :param max_tokens: The maximum number of tokens to generate in the completion. The token count of your prompt plus `max_tokens` cannot exceed the model's context length.

mistralai_gcp/fim.py

@@ -1,10 +1,9 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from .basesdk import BaseSDK
-from mistralai_gcp import models
+from mistralai_gcp import models, utils
 from mistralai_gcp._hooks import HookContext
-from mistralai_gcp.types import OptionalNullable, UNSET
-import mistralai_gcp.utils as utils
+from mistralai_gcp.types import Nullable, OptionalNullable, UNSET
 from mistralai_gcp.utils import eventstreaming
 from typing import Any, AsyncGenerator, Generator, Optional, Union
 
@@ -14,8 +13,8 @@ class Fim(BaseSDK):
     
     def stream(
         self, *,
+        model: Nullable[str],
         prompt: str,
-        model: OptionalNullable[str] = UNSET,
         temperature: Optional[float] = 0.7,
         top_p: Optional[float] = 1,
         max_tokens: OptionalNullable[int] = UNSET,
@@ -32,8 +31,8 @@ class Fim(BaseSDK):
 
         Mistral AI provides the ability to stream responses back to a client in order to allow partial results for certain requests. Tokens will be sent as data-only server-sent events as they become available, with the stream terminated by a data: [DONE] message. Otherwise, the server will hold the request open until the timeout or until completion, with the response containing the full result as JSON.
 
-        :param prompt: The text/code to complete.
         :param model: ID of the model to use. Only compatible for now with:   - `codestral-2405`   - `codestral-latest`
+        :param prompt: The text/code to complete.
         :param temperature: What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both.
         :param top_p: Nucleus sampling, where the model considers the results of the tokens with `top_p` probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.
         :param max_tokens: The maximum number of tokens to generate in the completion. The token count of your prompt plus `max_tokens` cannot exceed the model's context length.
@@ -121,8 +120,8 @@ class Fim(BaseSDK):
     
     async def stream_async(
         self, *,
+        model: Nullable[str],
         prompt: str,
-        model: OptionalNullable[str] = UNSET,
         temperature: Optional[float] = 0.7,
         top_p: Optional[float] = 1,
         max_tokens: OptionalNullable[int] = UNSET,
@@ -139,8 +138,8 @@ class Fim(BaseSDK):
 
         Mistral AI provides the ability to stream responses back to a client in order to allow partial results for certain requests. Tokens will be sent as data-only server-sent events as they become available, with the stream terminated by a data: [DONE] message. Otherwise, the server will hold the request open until the timeout or until completion, with the response containing the full result as JSON.
 
-        :param prompt: The text/code to complete.
         :param model: ID of the model to use. Only compatible for now with:   - `codestral-2405`   - `codestral-latest`
+        :param prompt: The text/code to complete.
         :param temperature: What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both.
         :param top_p: Nucleus sampling, where the model considers the results of the tokens with `top_p` probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.
         :param max_tokens: The maximum number of tokens to generate in the completion. The token count of your prompt plus `max_tokens` cannot exceed the model's context length.
@@ -226,10 +225,10 @@ class Fim(BaseSDK):
 
     
     
-    def create(
+    def complete(
         self, *,
+        model: Nullable[str],
         prompt: str,
-        model: OptionalNullable[str] = UNSET,
         temperature: Optional[float] = 0.7,
         top_p: Optional[float] = 1,
         max_tokens: OptionalNullable[int] = UNSET,
@@ -246,8 +245,8 @@ class Fim(BaseSDK):
 
         FIM completion.
 
-        :param prompt: The text/code to complete.
         :param model: ID of the model to use. Only compatible for now with:   - `codestral-2405`   - `codestral-latest`
+        :param prompt: The text/code to complete.
         :param temperature: What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both.
         :param top_p: Nucleus sampling, where the model considers the results of the tokens with `top_p` probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.
         :param max_tokens: The maximum number of tokens to generate in the completion. The token count of your prompt plus `max_tokens` cannot exceed the model's context length.
@@ -332,10 +331,10 @@ class Fim(BaseSDK):
 
     
     
-    async def create_async(
+    async def complete_async(
         self, *,
+        model: Nullable[str],
         prompt: str,
-        model: OptionalNullable[str] = UNSET,
         temperature: Optional[float] = 0.7,
         top_p: Optional[float] = 1,
         max_tokens: OptionalNullable[int] = UNSET,
@@ -352,8 +351,8 @@ class Fim(BaseSDK):
 
         FIM completion.
 
-        :param prompt: The text/code to complete.
         :param model: ID of the model to use. Only compatible for now with:   - `codestral-2405`   - `codestral-latest`
+        :param prompt: The text/code to complete.
         :param temperature: What sampling temperature to use, between 0.0 and 1.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or `top_p` but not both.
         :param top_p: Nucleus sampling, where the model considers the results of the tokens with `top_p` probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.
         :param max_tokens: The maximum number of tokens to generate in the completion. The token count of your prompt plus `max_tokens` cannot exceed the model's context length.

mistralai_gcp/models/__init__.py

@@ -1,7 +1,7 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from .assistantmessage import AssistantMessage, AssistantMessageRole, AssistantMessageTypedDict
-from .chatcompletionchoice import ChatCompletionChoice, ChatCompletionChoiceFinishReason, ChatCompletionChoiceTypedDict, Message, MessageTypedDict
+from .chatcompletionchoice import ChatCompletionChoice, ChatCompletionChoiceFinishReason, ChatCompletionChoiceTypedDict
 from .chatcompletionrequest import ChatCompletionRequest, ChatCompletionRequestMessages, ChatCompletionRequestMessagesTypedDict, ChatCompletionRequestStop, ChatCompletionRequestStopTypedDict, ChatCompletionRequestToolChoice, ChatCompletionRequestTypedDict
 from .chatcompletionresponse import ChatCompletionResponse, ChatCompletionResponseTypedDict
 from .chatcompletionstreamrequest import ChatCompletionStreamRequest, ChatCompletionStreamRequestTypedDict, Messages, MessagesTypedDict, Stop, StopTypedDict, ToolChoice
@@ -14,7 +14,7 @@ from .fimcompletionrequest import FIMCompletionRequest, FIMCompletionRequestStop
 from .fimcompletionresponse import FIMCompletionResponse, FIMCompletionResponseTypedDict
 from .fimcompletionstreamrequest import FIMCompletionStreamRequest, FIMCompletionStreamRequestStop, FIMCompletionStreamRequestStopTypedDict, FIMCompletionStreamRequestTypedDict
 from .function import Function, FunctionTypedDict
-from .functioncall import FunctionCall, FunctionCallTypedDict
+from .functioncall import Arguments, ArgumentsTypedDict, FunctionCall, FunctionCallTypedDict
 from .httpvalidationerror import HTTPValidationError, HTTPValidationErrorData
 from .responseformat import ResponseFormat, ResponseFormatTypedDict, ResponseFormats
 from .sdkerror import SDKError
@@ -28,4 +28,4 @@ from .usageinfo import UsageInfo, UsageInfoTypedDict
 from .usermessage import UserMessage, UserMessageContent, UserMessageContentTypedDict, UserMessageRole, UserMessageTypedDict
 from .validationerror import Loc, LocTypedDict, ValidationError, ValidationErrorTypedDict
 
-__all__ = ["AssistantMessage", "AssistantMessageRole", "AssistantMessageTypedDict", "ChatCompletionChoice", "ChatCompletionChoiceFinishReason", "ChatCompletionChoiceTypedDict", "ChatCompletionRequest", "ChatCompletionRequestMessages", "ChatCompletionRequestMessagesTypedDict", "ChatCompletionRequestStop", "ChatCompletionRequestStopTypedDict", "ChatCompletionRequestToolChoice", "ChatCompletionRequestTypedDict", "ChatCompletionResponse", "ChatCompletionResponseTypedDict", "ChatCompletionStreamRequest", "ChatCompletionStreamRequestTypedDict", "CompletionChunk", "CompletionChunkTypedDict", "CompletionEvent", "CompletionEventTypedDict", "CompletionResponseStreamChoice", "CompletionResponseStreamChoiceTypedDict", "Content", "ContentChunk", "ContentChunkTypedDict", "ContentTypedDict", "DeltaMessage", "DeltaMessageTypedDict", "FIMCompletionRequest", "FIMCompletionRequestStop", "FIMCompletionRequestStopTypedDict", "FIMCompletionRequestTypedDict", "FIMCompletionResponse", "FIMCompletionResponseTypedDict", "FIMCompletionStreamRequest", "FIMCompletionStreamRequestStop", "FIMCompletionStreamRequestStopTypedDict", "FIMCompletionStreamRequestTypedDict", "FinishReason", "Function", "FunctionCall", "FunctionCallTypedDict", "FunctionTypedDict", "HTTPValidationError", "HTTPValidationErrorData", "Loc", "LocTypedDict", "Message", "MessageTypedDict", "Messages", "MessagesTypedDict", "ResponseFormat", "ResponseFormatTypedDict", "ResponseFormats", "Role", "SDKError", "Security", "SecurityTypedDict", "Stop", "StopTypedDict", "SystemMessage", "SystemMessageTypedDict", "TextChunk", "TextChunkTypedDict", "Tool", "ToolCall", "ToolCallTypedDict", "ToolChoice", "ToolMessage", "ToolMessageRole", "ToolMessageTypedDict", "ToolTypedDict", "UsageInfo", "UsageInfoTypedDict", "UserMessage", "UserMessageContent", "UserMessageContentTypedDict", "UserMessageRole", "UserMessageTypedDict", "ValidationError", "ValidationErrorTypedDict"]
+__all__ = ["Arguments", "ArgumentsTypedDict", "AssistantMessage", "AssistantMessageRole", "AssistantMessageTypedDict", "ChatCompletionChoice", "ChatCompletionChoiceFinishReason", "ChatCompletionChoiceTypedDict", "ChatCompletionRequest", "ChatCompletionRequestMessages", "ChatCompletionRequestMessagesTypedDict", "ChatCompletionRequestStop", "ChatCompletionRequestStopTypedDict", "ChatCompletionRequestToolChoice", "ChatCompletionRequestTypedDict", "ChatCompletionResponse", "ChatCompletionResponseTypedDict", "ChatCompletionStreamRequest", "ChatCompletionStreamRequestTypedDict", "CompletionChunk", "CompletionChunkTypedDict", "CompletionEvent", "CompletionEventTypedDict", "CompletionResponseStreamChoice", "CompletionResponseStreamChoiceTypedDict", "Content", "ContentChunk", "ContentChunkTypedDict", "ContentTypedDict", "DeltaMessage", "DeltaMessageTypedDict", "FIMCompletionRequest", "FIMCompletionRequestStop", "FIMCompletionRequestStopTypedDict", "FIMCompletionRequestTypedDict", "FIMCompletionResponse", "FIMCompletionResponseTypedDict", "FIMCompletionStreamRequest", "FIMCompletionStreamRequestStop", "FIMCompletionStreamRequestStopTypedDict", "FIMCompletionStreamRequestTypedDict", "FinishReason", "Function", "FunctionCall", "FunctionCallTypedDict", "FunctionTypedDict", "HTTPValidationError", "HTTPValidationErrorData", "Loc", "LocTypedDict", "Messages", "MessagesTypedDict", "ResponseFormat", "ResponseFormatTypedDict", "ResponseFormats", "Role", "SDKError", "Security", "SecurityTypedDict", "Stop", "StopTypedDict", "SystemMessage", "SystemMessageTypedDict", "TextChunk", "TextChunkTypedDict", "Tool", "ToolCall", "ToolCallTypedDict", "ToolChoice", "ToolMessage", "ToolMessageRole", "ToolMessageTypedDict", "ToolTypedDict", "UsageInfo", "UsageInfoTypedDict", "UserMessage", "UserMessageContent", "UserMessageContentTypedDict", "UserMessageRole", "UserMessageTypedDict", "ValidationError", "ValidationErrorTypedDict"]

mistralai_gcp/models/assistantmessage.py

@@ -39,18 +39,13 @@ class AssistantMessage(BaseModel):
             k = f.alias or n
             val = serialized.get(k)
 
+            optional_nullable = k in optional_fields and k in nullable_fields
+            is_set = (self.__pydantic_fields_set__.intersection({n}) or k in null_default_fields) # pylint: disable=no-member
+
             if val is not None and val != UNSET_SENTINEL:
                 m[k] = val
             elif val != UNSET_SENTINEL and (
-                not k in optional_fields
-                or (
-                    k in optional_fields
-                    and k in nullable_fields
-                    and (
-                        self.__pydantic_fields_set__.intersection({n})
-                        or k in null_default_fields
-                    )  # pylint: disable=no-member
-                )
+                not k in optional_fields or (optional_nullable and is_set)
             ):
                 m[k] = val
 

mistralai_gcp/models/chatcompletionchoice.py

@@ -2,14 +2,9 @@
 
 from __future__ import annotations
 from .assistantmessage import AssistantMessage, AssistantMessageTypedDict
-from .systemmessage import SystemMessage, SystemMessageTypedDict
-from .toolmessage import ToolMessage, ToolMessageTypedDict
-from .usermessage import UserMessage, UserMessageTypedDict
 from mistralai_gcp.types import BaseModel
-from mistralai_gcp.utils import get_discriminator
-from pydantic import Discriminator, Tag
-from typing import Literal, Optional, TypedDict, Union
-from typing_extensions import Annotated, NotRequired
+from typing import Literal, Optional, TypedDict
+from typing_extensions import NotRequired
 
 
 ChatCompletionChoiceFinishReason = Literal["stop", "length", "model_length", "error", "tool_calls"]
@@ -17,17 +12,11 @@ ChatCompletionChoiceFinishReason = Literal["stop", "length", "model_length", "er
 class ChatCompletionChoiceTypedDict(TypedDict):
     index: int
     finish_reason: ChatCompletionChoiceFinishReason
-    message: NotRequired[MessageTypedDict]
+    message: NotRequired[AssistantMessageTypedDict]
     
 
 class ChatCompletionChoice(BaseModel):
     index: int
     finish_reason: ChatCompletionChoiceFinishReason
-    message: Optional[Message] = None
+    message: Optional[AssistantMessage] = None
     
-
-MessageTypedDict = Union[SystemMessageTypedDict, UserMessageTypedDict, AssistantMessageTypedDict, ToolMessageTypedDict]
-
-
-Message = Annotated[Union[Annotated[AssistantMessage, Tag("assistant")], Annotated[SystemMessage, Tag("system")], Annotated[ToolMessage, Tag("tool")], Annotated[UserMessage, Tag("user")]], Discriminator(lambda m: get_discriminator(m, "role", "role"))]
-