gllm-inference-binary 0.5.7__cp311-cp311-win_amd64.whl → 0.5.10b12__cp311-cp311-win_amd64.whl

gllm_inference/constants.pyi

@@ -7,3 +7,4 @@ INVOKER_PROPAGATED_MAX_RETRIES: int
 INVOKER_DEFAULT_TIMEOUT: float
 HEX_REPR_LENGTH: int
 HTTP_STATUS_CODE_PATTERNS: Incomplete
+SECONDS_TO_MILLISECONDS: int

gllm_inference/em_invoker/google_em_invoker.pyi

@@ -1,6 +1,6 @@
 from _typeshed import Incomplete
 from gllm_core.utils.retry import RetryConfig as RetryConfig
-from gllm_inference.constants import GOOGLE_SCOPES as GOOGLE_SCOPES
+from gllm_inference.constants import GOOGLE_SCOPES as GOOGLE_SCOPES, SECONDS_TO_MILLISECONDS as SECONDS_TO_MILLISECONDS
 from gllm_inference.em_invoker.em_invoker import BaseEMInvoker as BaseEMInvoker
 from gllm_inference.em_invoker.schema.google import Key as Key
 from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, Vector as Vector

gllm_inference/lm_invoker/anthropic_lm_invoker.pyi

@@ -1,11 +1,12 @@
 from _typeshed import Incomplete
 from gllm_core.event import EventEmitter as EventEmitter
+from gllm_core.schema.tool import Tool as Tool
 from gllm_core.utils.retry import RetryConfig as RetryConfig
 from gllm_inference.constants import INVOKER_PROPAGATED_MAX_RETRIES as INVOKER_PROPAGATED_MAX_RETRIES
 from gllm_inference.lm_invoker.lm_invoker import BaseLMInvoker as BaseLMInvoker
 from gllm_inference.lm_invoker.schema.anthropic import InputType as InputType, Key as Key, OutputType as OutputType
 from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, EmitDataType as EmitDataType, LMOutput as LMOutput, Message as Message, ModelId as ModelId, ModelProvider as ModelProvider, Reasoning as Reasoning, ResponseSchema as ResponseSchema, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
-from langchain_core.tools import Tool as Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -49,7 +50,7 @@ class AnthropicLMInvoker(BaseLMInvoker):
 
     Tool calling:
         Tool calling is a feature that allows the language model to call tools to perform tasks.
-        Tools can be passed to the via the `tools` parameter as a list of LangChain\'s `Tool` objects.
+        Tools can be passed to the via the `tools` parameter as a list of `Tool` objects.
         When tools are provided and the model decides to call a tool, the tool calls are stored in the
         `tool_calls` attribute in the output.
 
@@ -222,7 +223,7 @@ class AnthropicLMInvoker(BaseLMInvoker):
     client: Incomplete
     thinking: Incomplete
     thinking_budget: Incomplete
-    def __init__(self, model_name: str, api_key: str | None = None, model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, thinking: bool = False, thinking_budget: int = ...) -> None:
+    def __init__(self, model_name: str, api_key: str | None = None, model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, thinking: bool = False, thinking_budget: int = ...) -> None:
         """Initializes the AnthropicLmInvoker instance.
 
         Args:
@@ -232,7 +233,8 @@ class AnthropicLMInvoker(BaseLMInvoker):
             model_kwargs (dict[str, Any] | None, optional): Additional keyword arguments for the Anthropic client.
             default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the model.
                 Defaults to None.
-            tools (list[Tool] | None, optional): Tools provided to the model to enable tool calling. Defaults to None.
+            tools (list[Tool | LangChainTool] | None, optional): Tools provided to the model to enable tool calling.
+                Defaults to None, in which case an empty list is used.
             response_schema (ResponseSchema | None, optional): The schema of the response. If provided, the model will
                 output a structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema
                 dictionary. Defaults to None.
@@ -248,13 +250,13 @@ class AnthropicLMInvoker(BaseLMInvoker):
             1. `thinking` is True, but the `thinking_budget` is less than 1024.
             3. `response_schema` is provided, but `tools` or `thinking` are also provided.
         """
-    def set_tools(self, tools: list[Tool]) -> None:
+    def set_tools(self, tools: list[Tool | LangChainTool]) -> None:
         """Sets the tools for the Anthropic language model.
 
         This method sets the tools for the Anthropic language model. Any existing tools will be replaced.
 
         Args:
-            tools (list[Tool]): The list of tools to be used.
+            tools (list[Tool | LangChainTool]): The list of tools to be used.
 
         Raises:
             ValueError: If `response_schema` exists.

gllm_inference/lm_invoker/azure_openai_lm_invoker.pyi

@@ -1,10 +1,11 @@
 from _typeshed import Incomplete
+from gllm_core.schema.tool import Tool as Tool
 from gllm_core.utils.retry import RetryConfig as RetryConfig
 from gllm_inference.constants import DEFAULT_AZURE_OPENAI_API_VERSION as DEFAULT_AZURE_OPENAI_API_VERSION, INVOKER_PROPAGATED_MAX_RETRIES as INVOKER_PROPAGATED_MAX_RETRIES
 from gllm_inference.lm_invoker.openai_lm_invoker import OpenAILMInvoker as OpenAILMInvoker, ReasoningEffort as ReasoningEffort, ReasoningSummary as ReasoningSummary
 from gllm_inference.lm_invoker.schema.openai import Key as Key
 from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, ResponseSchema as ResponseSchema
-from langchain_core.tools import Tool as Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 class AzureOpenAILMInvoker(OpenAILMInvoker):
@@ -16,7 +17,7 @@ class AzureOpenAILMInvoker(OpenAILMInvoker):
         model_name (str): The name of the Azure OpenAI language model deployment.
         client (AsyncAzureOpenAI): The Azure OpenAI client instance.
         default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the model.
-        tools (list[Any]): The list of tools provided to the model to enable tool calling.
+        tools (list[Tool]): The list of tools provided to the model to enable tool calling.
         response_schema (ResponseSchema | None): The schema of the response. If provided, the model will output a
             structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema dictionary.
         output_analytics (bool): Whether to output the invocation analytics.
@@ -51,7 +52,7 @@ class AzureOpenAILMInvoker(OpenAILMInvoker):
 
     Tool calling:
         Tool calling is a feature that allows the language model to call tools to perform tasks.
-        Tools can be passed to the via the `tools` parameter as a list of LangChain\'s `Tool` objects.
+        Tools can be passed to the via the `tools` parameter as a list of `Tool` objects.
         When tools are provided and the model decides to call a tool, the tool calls are stored in the
         `tool_calls` attribute in the output.
 
@@ -219,7 +220,7 @@ class AzureOpenAILMInvoker(OpenAILMInvoker):
                 Defaults to an empty list.
     '''
     client: Incomplete
-    def __init__(self, azure_endpoint: str, azure_deployment: str, api_key: str | None = None, api_version: str = ..., model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, reasoning_effort: ReasoningEffort | None = None, reasoning_summary: ReasoningSummary | None = None) -> None:
+    def __init__(self, azure_endpoint: str, azure_deployment: str, api_key: str | None = None, api_version: str = ..., model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, reasoning_effort: ReasoningEffort | None = None, reasoning_summary: ReasoningSummary | None = None) -> None:
         """Initializes a new instance of the AzureOpenAILMInvoker class.
 
         Args:
@@ -232,7 +233,7 @@ class AzureOpenAILMInvoker(OpenAILMInvoker):
             model_kwargs (dict[str, Any] | None, optional): Additional model parameters. Defaults to None.
             default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the model.
                 Defaults to None.
-            tools (list[Tool] | None, optional): Tools provided to the language model to enable tool calling.
+            tools (list[Tool | LangChainTool] | None, optional): Tools provided to the model to enable tool calling.
                 Defaults to None.
             response_schema (ResponseSchema | None, optional): The schema of the response. If provided, the model will
                 output a structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema

gllm_inference/lm_invoker/bedrock_lm_invoker.pyi

@@ -1,10 +1,11 @@
 from _typeshed import Incomplete
 from gllm_core.event import EventEmitter as EventEmitter
+from gllm_core.schema.tool import Tool as Tool
 from gllm_core.utils.retry import RetryConfig as RetryConfig
 from gllm_inference.lm_invoker.lm_invoker import BaseLMInvoker as BaseLMInvoker
 from gllm_inference.lm_invoker.schema.bedrock import InputType as InputType, Key as Key, OutputType as OutputType
 from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, LMOutput as LMOutput, Message as Message, ModelId as ModelId, ModelProvider as ModelProvider, ResponseSchema as ResponseSchema, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
-from langchain_core.tools import Tool as Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -49,7 +50,7 @@ class BedrockLMInvoker(BaseLMInvoker):
 
     Tool calling:
         Tool calling is a feature that allows the language model to call tools to perform tasks.
-        Tools can be passed to the via the `tools` parameter as a list of LangChain\'s `Tool` objects.
+        Tools can be passed to the via the `tools` parameter as a list of `Tool` objects.
         When tools are provided and the model decides to call a tool, the tool calls are stored in the
         `tool_calls` attribute in the output.
 
@@ -179,7 +180,7 @@ class BedrockLMInvoker(BaseLMInvoker):
     '''
     session: Incomplete
     client_kwargs: Incomplete
-    def __init__(self, model_name: str, access_key_id: str | None = None, secret_access_key: str | None = None, region_name: str = 'us-east-1', model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None) -> None:
+    def __init__(self, model_name: str, access_key_id: str | None = None, secret_access_key: str | None = None, region_name: str = 'us-east-1', model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None) -> None:
         '''Initializes the BedrockLMInvoker instance.
 
         Args:
@@ -192,7 +193,8 @@ class BedrockLMInvoker(BaseLMInvoker):
             model_kwargs (dict[str, Any] | None, optional): Additional keyword arguments for the Bedrock client.
             default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the model.
                 Defaults to None.
-            tools (list[Tool] | None, optional): Tools provided to the model to enable tool calling. Defaults to None.
+            tools (list[Tool | LangChainTool] | None, optional): Tools provided to the model to enable tool calling.
+                Defaults to None.
             response_schema (ResponseSchema | None, optional): The schema of the response. If provided, the model will
                 output a structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema
                 dictionary. Defaults to None.
@@ -205,13 +207,13 @@ class BedrockLMInvoker(BaseLMInvoker):
             ValueError: If `access_key_id` or `secret_access_key` is neither provided nor set in the
                 `AWS_ACCESS_KEY_ID` or `AWS_SECRET_ACCESS_KEY` environment variables, respectively.
         '''
-    def set_tools(self, tools: list[Tool]) -> None:
+    def set_tools(self, tools: list[Tool | LangChainTool]) -> None:
         """Sets the tools for the Bedrock language model.
 
         This method sets the tools for the Bedrock language model. Any existing tools will be replaced.
 
         Args:
-            tools (list[Tool]): The list of tools to be used.
+            tools (list[Tool | LangChainTool]): The list of tools to be used.
 
         Raises:
             ValueError: If `response_schema` exists.

gllm_inference/lm_invoker/datasaur_lm_invoker.pyi

@@ -1,11 +1,12 @@
 from _typeshed import Incomplete
 from gllm_core.event import EventEmitter as EventEmitter
+from gllm_core.schema.tool import Tool as Tool
 from gllm_core.utils.retry import RetryConfig as RetryConfig
 from gllm_inference.constants import DOCUMENT_MIME_TYPES as DOCUMENT_MIME_TYPES, INVOKER_PROPAGATED_MAX_RETRIES as INVOKER_PROPAGATED_MAX_RETRIES
 from gllm_inference.lm_invoker.openai_compatible_lm_invoker import OpenAICompatibleLMInvoker as OpenAICompatibleLMInvoker
 from gllm_inference.lm_invoker.schema.datasaur import InputType as InputType, Key as Key
 from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, LMOutput as LMOutput, Message as Message, ModelId as ModelId, ModelProvider as ModelProvider, ResponseSchema as ResponseSchema, ToolCall as ToolCall, ToolResult as ToolResult
-from langchain_core.tools import Tool as Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -139,14 +140,14 @@ class DatasaurLMInvoker(OpenAICompatibleLMInvoker):
         Raises:
             ValueError: If the `api_key` is not provided and the `DATASAUR_API_KEY` environment variable is not set.
         """
-    def set_tools(self, tools: list[Tool]) -> None:
+    def set_tools(self, tools: list[Tool | LangChainTool]) -> None:
         """Sets the tools for the Datasaur LLM Projects Deployment API.
 
         This method is raises a `NotImplementedError` because the Datasaur LLM Projects Deployment API does not
         support tools.
 
         Args:
-            tools (list[Tool]): The list of tools to be used.
+            tools (list[Tool | LangChainTool]): The list of tools to be used.
 
         Raises:
             NotImplementedError: This method is not supported for the Datasaur LLM Projects Deployment API.

gllm_inference/lm_invoker/google_lm_invoker.pyi

@@ -1,11 +1,12 @@
 from _typeshed import Incomplete
 from gllm_core.event import EventEmitter as EventEmitter
+from gllm_core.schema.tool import Tool
 from gllm_core.utils.retry import RetryConfig as RetryConfig
-from gllm_inference.constants import GOOGLE_SCOPES as GOOGLE_SCOPES
+from gllm_inference.constants import GOOGLE_SCOPES as GOOGLE_SCOPES, SECONDS_TO_MILLISECONDS as SECONDS_TO_MILLISECONDS
 from gllm_inference.lm_invoker.lm_invoker import BaseLMInvoker as BaseLMInvoker
 from gllm_inference.lm_invoker.schema.google import InputType as InputType, Key as Key
 from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, EmitDataType as EmitDataType, LMOutput as LMOutput, Message as Message, MessageRole as MessageRole, ModelId as ModelId, ModelProvider as ModelProvider, Reasoning as Reasoning, ResponseSchema as ResponseSchema, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
-from langchain_core.tools import Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -79,7 +80,7 @@ class GoogleLMInvoker(BaseLMInvoker):
 
     Tool calling:
         Tool calling is a feature that allows the language model to call tools to perform tasks.
-        Tools can be passed to the via the `tools` parameter as a list of LangChain\'s `Tool` objects.
+        Tools can be passed to the via the `tools` parameter as a list of `Tool` objects.
         When tools are provided and the model decides to call a tool, the tool calls are stored in the
         `tool_calls` attribute in the output.
 
@@ -161,7 +162,12 @@ class GoogleLMInvoker(BaseLMInvoker):
         ```python
         LMOutput(
             response="Golden retriever is a good dog breed.",
-            token_usage=TokenUsage(input_tokens=100, output_tokens=50),
+            token_usage=TokenUsage(
+                input_tokens=1500,
+                output_tokens=200,
+                input_token_details=InputTokenDetails(cached_tokens=1200, uncached_tokens=300),
+                output_token_details=OutputTokenDetails(reasoning_tokens=180, response_tokens=20),
+            ),
             duration=0.729,
             finish_details={"finish_reason": "STOP", "finish_message": None},
         )
@@ -250,7 +256,7 @@ class GoogleLMInvoker(BaseLMInvoker):
     client_params: Incomplete
     thinking: Incomplete
     thinking_budget: Incomplete
-    def __init__(self, model_name: str, api_key: str | None = None, credentials_path: str | None = None, project_id: str | None = None, location: str = 'us-central1', model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, thinking: bool | None = None, thinking_budget: int = ...) -> None:
+    def __init__(self, model_name: str, api_key: str | None = None, credentials_path: str | None = None, project_id: str | None = None, location: str = 'us-central1', model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, thinking: bool | None = None, thinking_budget: int = ...) -> None:
         '''Initializes a new instance of the GoogleLMInvoker class.
 
         Args:
@@ -267,7 +273,7 @@ class GoogleLMInvoker(BaseLMInvoker):
                 client.
             default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the model.
                 Defaults to None.
-            tools (list[Tool] | None, optional): Tools provided to the language model to enable tool calling.
+            tools (list[Tool | LangChainTool] | None, optional): Tools provided to the model to enable tool calling.
                 Defaults to None.
             response_schema (ResponseSchema | None, optional): The schema of the response. If provided, the model will
                 output a structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema
@@ -284,13 +290,13 @@ class GoogleLMInvoker(BaseLMInvoker):
             If neither `api_key` nor `credentials_path` is provided, Google Gen AI will be used by default.
             The `GOOGLE_API_KEY` environment variable will be used for authentication.
         '''
-    def set_tools(self, tools: list[Tool]) -> None:
+    def set_tools(self, tools: list[Tool | LangChainTool]) -> None:
         """Sets the tools for the Google language model.
 
         This method sets the tools for the Google language model. Any existing tools will be replaced.
 
         Args:
-            tools (list[Tool]): The list of tools to be used.
+            tools (list[Tool | LangChainTool]): The list of tools to be used.
 
         Raises:
             ValueError: If `response_schema` exists.

gllm_inference/lm_invoker/langchain_lm_invoker.pyi

@@ -1,5 +1,6 @@
 from _typeshed import Incomplete
 from gllm_core.event import EventEmitter as EventEmitter
+from gllm_core.schema.tool import Tool as Tool
 from gllm_core.utils.retry import RetryConfig
 from gllm_inference.constants import INVOKER_DEFAULT_TIMEOUT as INVOKER_DEFAULT_TIMEOUT, INVOKER_PROPAGATED_MAX_RETRIES as INVOKER_PROPAGATED_MAX_RETRIES
 from gllm_inference.lm_invoker.lm_invoker import BaseLMInvoker as BaseLMInvoker
@@ -8,7 +9,7 @@ from gllm_inference.schema import Attachment as Attachment, AttachmentType as At
 from gllm_inference.utils import load_langchain_model as load_langchain_model, parse_model_data as parse_model_data
 from langchain_core.language_models import BaseChatModel as BaseChatModel
 from langchain_core.messages import BaseMessage as BaseMessage
-from langchain_core.tools import Tool as Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -77,7 +78,7 @@ class LangChainLMInvoker(BaseLMInvoker):
 
     Tool calling:
         Tool calling is a feature that allows the language model to call tools to perform tasks.
-        Tools can be passed to the via the `tools` parameter as a list of LangChain\'s `Tool` objects.
+        Tools can be passed to the via the `tools` parameter as a list of `Tool` objects.
         When tools are provided and the model decides to call a tool, the tool calls are stored in the
         `tool_calls` attribute in the output.
 
@@ -204,7 +205,7 @@ class LangChainLMInvoker(BaseLMInvoker):
                 Defaults to an empty list.
     '''
     model: Incomplete
-    def __init__(self, model: BaseChatModel | None = None, model_class_path: str | None = None, model_name: str | None = None, model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None) -> None:
+    def __init__(self, model: BaseChatModel | None = None, model_class_path: str | None = None, model_name: str | None = None, model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None) -> None:
         '''Initializes a new instance of the LangChainLMInvoker class.
 
         Args:
@@ -219,7 +220,8 @@ class LangChainLMInvoker(BaseLMInvoker):
                 `model_class_path` is provided. Defaults to None.
             default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the model.
                 Defaults to None.
-            tools (list[Tool] | None, optional): Tools provided to the model to enable tool calling. Defaults to None.
+            tools (list[Tool | LangChainTool] | None, optional): Tools provided to the model to enable tool calling.
+                Defaults to None.
             response_schema (ResponseSchema | None, optional): The schema of the response. If provided, the model will
                 output a structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema
                 dictionary. Defaults to None.
@@ -231,7 +233,7 @@ class LangChainLMInvoker(BaseLMInvoker):
             ValueError: If `response_schema` is provided, but `tools` are also provided.
         '''
     tools: Incomplete
-    def set_tools(self, tools: list[Tool]) -> None:
+    def set_tools(self, tools: list[Tool | LangChainTool]) -> None:
         """Sets the tools for LangChain's BaseChatModel.
 
         This method sets the tools for LangChain's BaseChatModel. Any existing tools will be replaced.

gllm_inference/lm_invoker/litellm_lm_invoker.pyi

@@ -1,10 +1,11 @@
 from _typeshed import Incomplete
 from gllm_core.event import EventEmitter as EventEmitter
+from gllm_core.schema.tool import Tool as Tool
 from gllm_core.utils.retry import RetryConfig as RetryConfig
 from gllm_inference.lm_invoker.openai_compatible_lm_invoker import OpenAICompatibleLMInvoker as OpenAICompatibleLMInvoker
 from gllm_inference.lm_invoker.openai_lm_invoker import ReasoningEffort as ReasoningEffort
 from gllm_inference.schema import AttachmentType as AttachmentType, LMOutput as LMOutput, ModelId as ModelId, ModelProvider as ModelProvider, ResponseSchema as ResponseSchema
-from langchain_core.tools import Tool as Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -58,7 +59,7 @@ class LiteLLMLMInvoker(OpenAICompatibleLMInvoker):
 
     Tool calling:
         Tool calling is a feature that allows the language model to call tools to perform tasks.
-        Tools can be passed to the via the `tools` parameter as a list of LangChain\'s `Tool` objects.
+        Tools can be passed to the via the `tools` parameter as a list of `Tool` objects.
         When tools are provided and the model decides to call a tool, the tool calls are stored in the
         `tool_calls` attribute in the output.
 
@@ -227,14 +228,15 @@ class LiteLLMLMInvoker(OpenAICompatibleLMInvoker):
                 Defaults to an empty list.
     '''
     completion: Incomplete
-    def __init__(self, model_id: str, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, reasoning_effort: ReasoningEffort | None = None) -> None:
+    def __init__(self, model_id: str, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, reasoning_effort: ReasoningEffort | None = None) -> None:
         """Initializes a new instance of the LiteLLMLMInvoker class.
 
         Args:
             model_id (str): The ID of the model to use. Must be in the format of `provider/model_name`.
             default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the model.
                 Defaults to None.
-            tools (list[Tool] | None, optional): Tools provided to the model to enable tool calling. Defaults to None.
+            tools (list[Tool | LangChainTool] | None, optional): Tools provided to the model to enable tool calling.
+                Defaults to None.
             response_schema (ResponseSchema | None, optional): The schema of the response. If provided, the model will
                 output a structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema
                 dictionary. Defaults to None.

gllm_inference/lm_invoker/lm_invoker.pyi

@@ -2,11 +2,12 @@ import abc
 from _typeshed import Incomplete
 from abc import ABC
 from gllm_core.event import EventEmitter as EventEmitter
-from gllm_core.utils.retry import RetryConfig
+from gllm_core.schema.tool import Tool
+from gllm_core.utils import RetryConfig
 from gllm_inference.constants import DOCUMENT_MIME_TYPES as DOCUMENT_MIME_TYPES, INVOKER_DEFAULT_TIMEOUT as INVOKER_DEFAULT_TIMEOUT
 from gllm_inference.exceptions import parse_error_message as parse_error_message
 from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, EmitDataType as EmitDataType, LMOutput as LMOutput, Message as Message, MessageContent as MessageContent, MessageRole as MessageRole, ModelId as ModelId, Reasoning as Reasoning, ResponseSchema as ResponseSchema, ToolCall as ToolCall, ToolResult as ToolResult
-from langchain_core.tools import Tool as Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 class _Key:
@@ -16,8 +17,14 @@ class _Key:
     DATA_TYPE: str
     DATA_VALUE: str
     DEFAULT: str
+    DESCRIPTION: str
+    FUNCTION: str
+    META: str
+    NAME: str
+    PARAMETERS: str
     PROPERTIES: str
     REQUIRED: str
+    TITLE: str
     TYPE: str
 
 class _InputType:
@@ -46,7 +53,7 @@ class BaseLMInvoker(ABC, metaclass=abc.ABCMeta):
     response_schema: Incomplete
     output_analytics: Incomplete
     retry_config: Incomplete
-    def __init__(self, model_id: ModelId, default_hyperparameters: dict[str, Any] | None = None, supported_attachments: set[str] | None = None, tools: list[Tool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None) -> None:
+    def __init__(self, model_id: ModelId, default_hyperparameters: dict[str, Any] | None = None, supported_attachments: set[str] | None = None, tools: list[Tool | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None) -> None:
         """Initializes a new instance of the BaseLMInvoker class.
 
         Args:
@@ -55,7 +62,7 @@ class BaseLMInvoker(ABC, metaclass=abc.ABCMeta):
                 language model. Defaults to None, in which case an empty dictionary is used.
             supported_attachments (set[str] | None, optional): A set of supported attachment types. Defaults to None,
                 in which case an empty set is used (indicating that no attachments are supported).
-            tools (list[Tool] | None, optional): Tools provided to the language model to enable tool calling.
+            tools (list[Tool | LangChainTool] | None, optional): Tools provided to the model to enable tool calling.
                 Defaults to None, in which case an empty list is used.
             response_schema (ResponseSchema | None, optional): The schema of the response. If provided, the model will
                 output a structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema
@@ -85,13 +92,13 @@ class BaseLMInvoker(ABC, metaclass=abc.ABCMeta):
         Returns:
             str: The name of the language model.
         """
-    def set_tools(self, tools: list[Tool]) -> None:
+    def set_tools(self, tools: list[Tool | LangChainTool]) -> None:
         """Sets the tools for the language model.
 
         This method sets the tools for the language model. Any existing tools will be replaced.
 
         Args:
-            tools (list[Tool]): The list of tools to be used.
+            tools (list[Tool | LangChainTool]): The list of tools to be used.
         """
     def clear_tools(self) -> None:
         """Clears the tools for the language model.

gllm_inference/lm_invoker/openai_compatible_lm_invoker.pyi

@@ -1,12 +1,13 @@
 from _typeshed import Incomplete
 from gllm_core.event import EventEmitter as EventEmitter
+from gllm_core.schema.tool import Tool as Tool
 from gllm_core.utils.retry import RetryConfig as RetryConfig
 from gllm_inference.constants import INVOKER_PROPAGATED_MAX_RETRIES as INVOKER_PROPAGATED_MAX_RETRIES
 from gllm_inference.lm_invoker.lm_invoker import BaseLMInvoker as BaseLMInvoker
 from gllm_inference.lm_invoker.schema.openai_compatible import InputType as InputType, Key as Key, ReasoningEffort as ReasoningEffort
 from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, EmitDataType as EmitDataType, LMOutput as LMOutput, Message as Message, MessageRole as MessageRole, ModelId as ModelId, ModelProvider as ModelProvider, Reasoning as Reasoning, ResponseSchema as ResponseSchema, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
 from gllm_inference.utils import validate_string_enum as validate_string_enum
-from langchain_core.tools import Tool as Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -20,7 +21,7 @@ class OpenAICompatibleLMInvoker(BaseLMInvoker):
         model_name (str): The name of the language model.
         client (AsyncOpenAI): The OpenAI client instance.
         default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the model.
-        tools (list[Any]): The list of tools provided to the model to enable tool calling.
+        tools (list[Tool]): The list of tools provided to the model to enable tool calling.
         response_schema (ResponseSchema | None): The schema of the response. If provided, the model will output a
             structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema dictionary.
         output_analytics (bool): Whether to output the invocation analytics.
@@ -63,7 +64,7 @@ class OpenAICompatibleLMInvoker(BaseLMInvoker):
 
     Tool calling:
         Tool calling is a feature that allows the language model to call tools to perform tasks.
-        Tools can be passed to the via the `tools` parameter as a list of LangChain\'s `Tool` objects.
+        Tools can be passed to the via the `tools` parameter as a list of `Tool` objects.
         When tools are provided and the model decides to call a tool, the tool calls are stored in the
         `tool_calls` attribute in the output.
 
@@ -231,7 +232,7 @@ class OpenAICompatibleLMInvoker(BaseLMInvoker):
                 Defaults to an empty list.
     '''
     client: Incomplete
-    def __init__(self, model_name: str, base_url: str, api_key: str | None = None, model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, reasoning_effort: ReasoningEffort | None = None) -> None:
+    def __init__(self, model_name: str, base_url: str, api_key: str | None = None, model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, reasoning_effort: ReasoningEffort | None = None) -> None:
         '''Initializes a new instance of the OpenAICompatibleLMInvoker class.
 
         Args:
@@ -243,7 +244,7 @@ class OpenAICompatibleLMInvoker(BaseLMInvoker):
             model_kwargs (dict[str, Any] | None, optional): Additional model parameters. Defaults to None.
             default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the model.
                 Defaults to None.
-            tools (list[Tool] | None, optional): Tools provided to the language model to enable tool calling.
+            tools (list[Tool | LangChainTool] | None, optional): Tools provided to the model to enable tool calling.
                 Defaults to None.
             response_schema (ResponseSchema | None, optional): The schema of the response. If provided, the model will
                 output a structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema

gllm_inference/lm_invoker/openai_lm_invoker.pyi

@@ -1,12 +1,13 @@
 from _typeshed import Incomplete
 from gllm_core.event import EventEmitter as EventEmitter
+from gllm_core.schema.tool import Tool as Tool
 from gllm_core.utils.retry import RetryConfig as RetryConfig
 from gllm_inference.constants import INVOKER_PROPAGATED_MAX_RETRIES as INVOKER_PROPAGATED_MAX_RETRIES
 from gllm_inference.lm_invoker.lm_invoker import BaseLMInvoker as BaseLMInvoker
 from gllm_inference.lm_invoker.schema.openai import InputType as InputType, Key as Key, OutputType as OutputType, ReasoningEffort as ReasoningEffort, ReasoningSummary as ReasoningSummary
 from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, CodeExecResult as CodeExecResult, EmitDataType as EmitDataType, LMOutput as LMOutput, Message as Message, MessageRole as MessageRole, ModelId as ModelId, ModelProvider as ModelProvider, Reasoning as Reasoning, ResponseSchema as ResponseSchema, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
 from gllm_inference.utils import validate_string_enum as validate_string_enum
-from langchain_core.tools import Tool as Tool
+from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -20,7 +21,7 @@ class OpenAILMInvoker(BaseLMInvoker):
         model_name (str): The name of the language model.
         client (AsyncOpenAI): The OpenAI client instance.
         default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the model.
-        tools (list[Any]): The list of tools provided to the model to enable tool calling.
+        tools (list[Tool]): The list of tools provided to the model to enable tool calling.
         response_schema (ResponseSchema | None): The schema of the response. If provided, the model will output a
             structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema dictionary.
         output_analytics (bool): Whether to output the invocation analytics.
@@ -52,7 +53,7 @@ class OpenAILMInvoker(BaseLMInvoker):
 
     Tool calling:
         Tool calling is a feature that allows the language model to call tools to perform tasks.
-        Tools can be passed to the via the `tools` parameter as a list of LangChain\'s `Tool` objects.
+        Tools can be passed to the via the `tools` parameter as a list of `Tool` objects.
         When tools are provided and the model decides to call a tool, the tool calls are stored in the
         `tool_calls` attribute in the output.
 
@@ -319,7 +320,7 @@ class OpenAILMInvoker(BaseLMInvoker):
                 enabled and the language model decides to execute any codes. Defaults to an empty list.
     '''
     client: Incomplete
-    def __init__(self, model_name: str, api_key: str | None = None, model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, reasoning_effort: ReasoningEffort | None = None, reasoning_summary: ReasoningSummary | None = None, code_interpreter: bool = False, web_search: bool = False) -> None:
+    def __init__(self, model_name: str, api_key: str | None = None, model_kwargs: dict[str, Any] | None = None, default_hyperparameters: dict[str, Any] | None = None, tools: list[Tool | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None, reasoning_effort: ReasoningEffort | None = None, reasoning_summary: ReasoningSummary | None = None, code_interpreter: bool = False, web_search: bool = False) -> None:
         """Initializes a new instance of the OpenAILMInvoker class.
 
         Args:
@@ -329,8 +330,8 @@ class OpenAILMInvoker(BaseLMInvoker):
             model_kwargs (dict[str, Any] | None, optional): Additional model parameters. Defaults to None.
             default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the model.
                 Defaults to None.
-            tools (list[Tool] | None, optional): Tools provided to the language model to enable tool calling.
-                Defaults to None.
+            tools (list[Tool | LangChainTool] | None, optional): Tools provided to the model to enable tool calling.
+                Defaults to None, in which case an empty list is used.
             response_schema (ResponseSchema | None, optional): The schema of the response. If provided, the model will
                 output a structured response as defined by the schema. Supports both Pydantic BaseModel and JSON schema
                 dictionary. Defaults to None.

gllm_inference/lm_invoker/schema/anthropic.pyi

@@ -4,7 +4,6 @@ class Key:
     CONTENT: str
     DATA: str
     DESCRIPTION: str
-    FUNCTION: str
     ID: str
     INPUT: str
     INPUT_SCHEMA: str

gllm_inference/lm_invoker/schema/openai.pyi

@@ -8,11 +8,11 @@ class Key:
     CONTENT: str
     DEFAULT: str
     DEFS: str
+    DESCRIPTION: str
     EFFORT: str
     FILE_DATA: str
     FILENAME: str
     FORMAT: str
-    FUNCTION: str
     ID: str
     IMAGE_URL: str
     INCLUDE: str

gllm_inference/lm_invoker/schema/openai_compatible.pyi

@@ -7,6 +7,7 @@ class Key:
     CHOICES: str
     DATA: str
     DEFS: str
+    DESCRIPTION: str
     EFFORT: str
     FINISH_REASON: str
     FORMAT: str
@@ -18,6 +19,7 @@ class Key:
     MAX_RETRIES: str
     MESSAGE: str
     NAME: str
+    PARAMETERS: str
     RESPONSE_FORMAT: str
     ROLE: str
     SCHEMA: str

gllm_inference/schema/__init__.pyi

@@ -5,9 +5,9 @@ from gllm_inference.schema.lm_output import LMOutput as LMOutput
 from gllm_inference.schema.message import Message as Message
 from gllm_inference.schema.model_id import ModelId as ModelId, ModelProvider as ModelProvider
 from gllm_inference.schema.reasoning import Reasoning as Reasoning
-from gllm_inference.schema.token_usage import TokenUsage as TokenUsage
+from gllm_inference.schema.token_usage import InputTokenDetails as InputTokenDetails, OutputTokenDetails as OutputTokenDetails, TokenUsage as TokenUsage
 from gllm_inference.schema.tool_call import ToolCall as ToolCall
 from gllm_inference.schema.tool_result import ToolResult as ToolResult
 from gllm_inference.schema.type_alias import EMContent as EMContent, ErrorResponse as ErrorResponse, MessageContent as MessageContent, ResponseSchema as ResponseSchema, Vector as Vector
 
-__all__ = ['Attachment', 'AttachmentType', 'CodeExecResult', 'EMContent', 'EmitDataType', 'ErrorResponse', 'MessageContent', 'LMOutput', 'ModelId', 'ModelProvider', 'Message', 'MessageRole', 'Reasoning', 'ResponseSchema', 'TokenUsage', 'ToolCall', 'ToolResult', 'Vector']
+__all__ = ['Attachment', 'AttachmentType', 'CodeExecResult', 'EMContent', 'EmitDataType', 'ErrorResponse', 'InputTokenDetails', 'MessageContent', 'LMOutput', 'ModelId', 'ModelProvider', 'Message', 'MessageRole', 'OutputTokenDetails', 'Reasoning', 'ResponseSchema', 'TokenUsage', 'ToolCall', 'ToolResult', 'Vector']

gllm_inference/schema/token_usage.pyi

@@ -1,11 +1,75 @@
 from pydantic import BaseModel
 
+class InputTokenDetails(BaseModel):
+    """Defines the input token details schema.
+
+    Attributes:
+        cached_tokens (int): The number of cached tokens. Defaults to 0.
+        uncached_tokens (int): The number of uncached tokens. Defaults to 0.
+    """
+    cached_tokens: int
+    uncached_tokens: int
+    def __add__(self, other: InputTokenDetails) -> InputTokenDetails:
+        """Add two InputTokenDetails objects together.
+
+        Args:
+            other (InputTokenDetails): The other InputTokenDetails object to add.
+
+        Returns:
+            InputTokenDetails: A new InputTokenDetails object with summed values.
+        """
+
+class OutputTokenDetails(BaseModel):
+    """Defines the output token details schema.
+
+    Attributes:
+        reasoning_tokens (int): The number of reasoning tokens. Defaults to 0.
+        response_tokens (int): The number of response tokens. Defaults to 0.
+    """
+    reasoning_tokens: int
+    response_tokens: int
+    def __add__(self, other: OutputTokenDetails) -> OutputTokenDetails:
+        """Add two OutputTokenDetails objects together.
+
+        Args:
+            other (OutputTokenDetails): The other OutputTokenDetails object to add.
+
+        Returns:
+            OutputTokenDetails: A new OutputTokenDetails object with summed values.
+        """
+
 class TokenUsage(BaseModel):
     """Defines the token usage data structure of a language model.
 
     Attributes:
-        input_tokens (int): The number of input tokens.
-        output_tokens (int): The number of output tokens.
+        input_tokens (int): The number of input tokens. Defaults to 0.
+        output_tokens (int): The number of output tokens. Defaults to 0.
+        input_token_details (InputTokenDetails | None): The details of the input tokens. Defaults to None.
+        output_token_details (OutputTokenDetails | None): The details of the output tokens. Defaults to None.
     """
     input_tokens: int
     output_tokens: int
+    input_token_details: InputTokenDetails | None
+    output_token_details: OutputTokenDetails | None
+    @classmethod
+    def from_token_details(cls, input_tokens: int | None = None, output_tokens: int | None = None, cached_tokens: int | None = None, reasoning_tokens: int | None = None) -> TokenUsage:
+        """Creates a TokenUsage from token details.
+
+        Args:
+            input_tokens (int | None): The number of input tokens. Defaults to None.
+            output_tokens (int | None): The number of output tokens. Defaults to None.
+            cached_tokens (int | None): The number of cached tokens. Defaults to None.
+            reasoning_tokens (int | None): The number of reasoning tokens. Defaults to None.
+
+        Returns:
+            TokenUsage: The instantiated TokenUsage.
+        """
+    def __add__(self, other: TokenUsage) -> TokenUsage:
+        """Add two TokenUsage objects together.
+
+        Args:
+            other (TokenUsage): The other TokenUsage object to add.
+
+        Returns:
+            TokenUsage: A new TokenUsage object with summed values.
+        """

gllm_inference.pyi

@@ -75,9 +75,9 @@ import requests
 import gllm_inference.schema.ErrorResponse
 import gllm_core.constants
 import gllm_core.event
+import gllm_core.schema
+import gllm_core.schema.tool
 import langchain_core.tools
-import langchain_core.utils
-import langchain_core.utils.function_calling
 import gllm_inference.schema.EmitDataType
 import gllm_inference.schema.LMOutput
 import gllm_inference.schema.Message
@@ -88,13 +88,14 @@ import gllm_inference.schema.ToolCall
 import gllm_inference.schema.ToolResult
 import anthropic
 import aioboto3
-import gllm_core.schema
 import gllm_inference.schema.MessageRole
 import langchain_core.language_models
 import langchain_core.messages
 import litellm
 import time
 import jsonschema
+import langchain_core.utils
+import langchain_core.utils.function_calling
 import gllm_inference.schema.MessageContent
 import gllm_inference.utils.validate_string_enum
 import gllm_inference.schema.CodeExecResult

{gllm_inference_binary-0.5.7.dist-info → gllm_inference_binary-0.5.10b12.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gllm-inference-binary
-Version: 0.5.7
+Version: 0.5.10b12
 Summary: A library containing components related to model inferences in Gen AI applications.
 Author: Henry Wicaksono
 Author-email: henry.wicaksono@gdplabs.id

{gllm_inference_binary-0.5.7.dist-info → gllm_inference_binary-0.5.10b12.dist-info}/RECORD

@@ -8,11 +8,11 @@ gllm_inference/catalog/__init__.pyi,sha256=HWgPKWIzprpMHRKe_qN9BZSIQhVhrqiyjLjIX
 gllm_inference/catalog/catalog.pyi,sha256=eWPqgQKi-SJGHabi_XOTEKpAj96OSRypKsb5ZEC1VWU,4911
 gllm_inference/catalog/lm_request_processor_catalog.pyi,sha256=GemCEjFRHNChtNOfbyXSVsJiA3klOCAe_X11fnymhYs,5540
 gllm_inference/catalog/prompt_builder_catalog.pyi,sha256=iViWB4SaezzjQY4UY1YxeoXUNxqxa2cTJGaD9JSx4Q8,3279
-gllm_inference/constants.pyi,sha256=xSET67ZCfeVK4a2ji1FZyQxs5DUTIsN7S6H7-F-ewZ0,265
+gllm_inference/constants.pyi,sha256=q1bzWw4fcz2RbdqQs3WYpvWc5aeassX-uQ5F0pDcKw4,295
 gllm_inference/em_invoker/__init__.pyi,sha256=eZifmg3ZS3YdFUwbGPTurrfF4oV_MAPvqErJe7oTpZI,882
 gllm_inference/em_invoker/azure_openai_em_invoker.pyi,sha256=QimqPII-KN9OgsfH1Iubn_tCHhtWjPQ5rilZoT6Ir-U,4688
 gllm_inference/em_invoker/em_invoker.pyi,sha256=KX4i0xBWR5j6z14nEL6T8at3StKfdf3miQ4xixtYhZk,4424
-gllm_inference/em_invoker/google_em_invoker.pyi,sha256=YJtRJs7bNGNEfTKtj3IVP1XkLcJ3LRmcAC80zzOHxKw,6254
+gllm_inference/em_invoker/google_em_invoker.pyi,sha256=c0XJuLS4ji-Y9aHmbb8PRLwrbH6QA2WtdS1-2HbWFq8,6306
 gllm_inference/em_invoker/langchain/__init__.pyi,sha256=aOTlRvS9aG1tBErjsmhe75s4Sq-g2z9ArfGqNW7QyEs,151
 gllm_inference/em_invoker/langchain/em_invoker_embeddings.pyi,sha256=gEX21gJLngUh9fZo8v6Vbh0gpWFFqS2S-dGNZSrDjFQ,2409
 gllm_inference/em_invoker/langchain_em_invoker.pyi,sha256=sFmsRE89MIdnD8g0VSMsdLvtfZL6dfPkUtDhH_WfgLc,2823
@@ -31,24 +31,24 @@ gllm_inference/exceptions/__init__.pyi,sha256=2F05RytXZIKaOJScb1pD0O0bATIQHVeEAY
 gllm_inference/exceptions/error_parser.pyi,sha256=ggmh8DJXdwFJInNLrP24WVJt_4raxbAVxzXRQgBpndA,2441
 gllm_inference/exceptions/exceptions.pyi,sha256=ViXvIzm7tLcstjqfwC6nPziDg0UAmoUAWZVWrAJyp3w,4763
 gllm_inference/lm_invoker/__init__.pyi,sha256=g-wu6W6ly_WAVPLDWKjt4J5cMo-CJ1x5unuObVSUnug,1115
-gllm_inference/lm_invoker/anthropic_lm_invoker.pyi,sha256=Jpu7YPr_oUB4WurY3MqDT9J-PgDDRakj7mtrle8S3gE,15038
-gllm_inference/lm_invoker/azure_openai_lm_invoker.pyi,sha256=h71Redwp4a4Ibq3tVTtwM7FUX4AqIlmty4i0Dt8lglk,14788
-gllm_inference/lm_invoker/bedrock_lm_invoker.pyi,sha256=uOppVYAy2G7TnIK_BsRllW0akP3x14zNjrfwVrTSo8I,12530
-gllm_inference/lm_invoker/datasaur_lm_invoker.pyi,sha256=AzjC-zEJt73I9wtgckZAlrlsC97U86cLksbYzO1XDK0,9280
-gllm_inference/lm_invoker/google_lm_invoker.pyi,sha256=a5ALWfhv8zW3go4gC4G-dGjQHbpQwHbSr6wdLiKiCKk,16702
-gllm_inference/lm_invoker/langchain_lm_invoker.pyi,sha256=iD_nFw0UGX9AHSEZT57T6e2ft2ApHW4sOKk49BDpBO8,13413
-gllm_inference/lm_invoker/litellm_lm_invoker.pyi,sha256=ye73iH8wXUXzAd87JCjup5wmgVxTNhgbebRNh1hCtjE,13283
-gllm_inference/lm_invoker/lm_invoker.pyi,sha256=1wH81ssLRLLSGdf7CMQ5CZqKLcuJZHmElwBjQwitqfg,7754
-gllm_inference/lm_invoker/openai_compatible_lm_invoker.pyi,sha256=jgeRQPw9IH24NoZ9dN8aW2w29D7Unn4w5L-6dhFX6l0,15125
-gllm_inference/lm_invoker/openai_lm_invoker.pyi,sha256=oXK_-nk3sdFnYIEPGkDCg4UWNyRvvMvVjkj7jozZhRk,19787
+gllm_inference/lm_invoker/anthropic_lm_invoker.pyi,sha256=85uvShLv4-eiGOpTMgwWpQGZXPW6XaB6GrexBmxg_sQ,15200
+gllm_inference/lm_invoker/azure_openai_lm_invoker.pyi,sha256=N2TjGz5Gi6xiLkAgI6SzWq_V3tj66HJfMNff7d04uU0,14856
+gllm_inference/lm_invoker/bedrock_lm_invoker.pyi,sha256=ae5P_9sjtcOgMIUaRchvp8F0FujoeP4e2F_OoHSe_go,12655
+gllm_inference/lm_invoker/datasaur_lm_invoker.pyi,sha256=c4H3TOz0LIhWjokCCdQ4asiwQR4_LPyaimo4RAqU9es,9369
+gllm_inference/lm_invoker/google_lm_invoker.pyi,sha256=I3plg_oVuTl0hiShFBmCYPclP4gWbzU61xUSgon24Ew,17102
+gllm_inference/lm_invoker/langchain_lm_invoker.pyi,sha256=bBGOxJfjnzOtDR4kH4PuCiOCKEPu8rTqzZodTXCHQ2k,13522
+gllm_inference/lm_invoker/litellm_lm_invoker.pyi,sha256=HHwW7i8ryXHI23JZQwscyva6aPmPOB13Muhf7gaaMUM,13376
+gllm_inference/lm_invoker/lm_invoker.pyi,sha256=YNJ0Sh_BOl1WbC69xvuxWM75qyByXJSXAYWSwtQ84cc,7960
+gllm_inference/lm_invoker/openai_compatible_lm_invoker.pyi,sha256=U9dolHJT1pDsiiyrdpSAAdcBkil4_qeG_3BKfygq8GM,15193
+gllm_inference/lm_invoker/openai_lm_invoker.pyi,sha256=SEHWAwpT8KmIQukurXtXOU2xyU2rp_HtM2SARsBF3dU,19892
 gllm_inference/lm_invoker/schema/__init__.pyi,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-gllm_inference/lm_invoker/schema/anthropic.pyi,sha256=LWhE6th4dpZwk7As11p_rhhvVoPXBsF3oAy5CtTMBrg,1060
+gllm_inference/lm_invoker/schema/anthropic.pyi,sha256=lGJ7xYLchdtv6003Is4GcaKiGdbmIOAzQsaldKG0Aww,1041
 gllm_inference/lm_invoker/schema/bedrock.pyi,sha256=H3attoGWhBA725W4FpXw7Mty46N9jHKjw9PT-0lMEJs,975
 gllm_inference/lm_invoker/schema/datasaur.pyi,sha256=GLv6XAwKtWyRrX6EsbEufYjkPffHNiEpXwJOn9HqxMA,242
 gllm_inference/lm_invoker/schema/google.pyi,sha256=elXHrUMS46pbTsulk7hBXVVFcT022iD-_U_I590xeV8,529
 gllm_inference/lm_invoker/schema/langchain.pyi,sha256=uEG0DSD0z4L_rDMkBm-TtUy5oTyEHEEJWiLsYvFf1sw,431
-gllm_inference/lm_invoker/schema/openai.pyi,sha256=CNkIGljwRyQYx0krONX1ik9hwBiN45t9vBk-ZY45rP4,1989
-gllm_inference/lm_invoker/schema/openai_compatible.pyi,sha256=2KXyTy43SHz0K_hB0WGXVuaHXL7bwYFIzWHZ_jmdDUo,1114
+gllm_inference/lm_invoker/schema/openai.pyi,sha256=Cxp5QMkF6lspcVUgCNZR1qDK43Fj6OoEdOiQ1x5arsQ,1992
+gllm_inference/lm_invoker/schema/openai_compatible.pyi,sha256=FnRfc3CiqY-y6WmZqi2OhxOnNrZENBEXCmk2WPADkBQ,1157
 gllm_inference/model/__init__.pyi,sha256=JKQB0wVSVYD-_tdRkG7N_oEVAKGCcoBw0BUOUMLieFo,602
 gllm_inference/model/em/__init__.pyi,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 gllm_inference/model/em/google_em.pyi,sha256=c53H-KNdNOK9ppPLyOSkmCA890eF5FsMd05upkPIzF0,487
@@ -74,7 +74,7 @@ gllm_inference/prompt_formatter/prompt_formatter.pyi,sha256=hAc6rxWc6JSYdD-OypLi
 gllm_inference/request_processor/__init__.pyi,sha256=giEme2WFQhgyKiBZHhSet0_nKSCHwGy-_2p6NRzg0Zc,231
 gllm_inference/request_processor/lm_request_processor.pyi,sha256=rInXhC95BvQnw9q98KZWpjPH8Q_TV4zC2ycNjypEBZ4,5516
 gllm_inference/request_processor/uses_lm_mixin.pyi,sha256=znBG4AWWm_H70Qqrc1mO4ohmWotX9id81Fqe-x9Qa6Q,2371
-gllm_inference/schema/__init__.pyi,sha256=-ldt0xJQJirVNdwLFev3bmzmFRw9HSUWBRmmIVH7uyU,1251
+gllm_inference/schema/__init__.pyi,sha256=Mg9aKyvShNaB4XmqLWcZZ0arSNJhT2g1hhIqP1IBuaM,1376
 gllm_inference/schema/attachment.pyi,sha256=9zgAjGXBjLfzPGaKi68FMW6b5mXdEA352nDe-ynOSvY,3385
 gllm_inference/schema/code_exec_result.pyi,sha256=WQ-ARoGM9r6nyRX-A0Ro1XKiqrc9R3jRYXZpu_xo5S4,573
 gllm_inference/schema/enums.pyi,sha256=SQ9mXt8j7uK333uUnUHRs-mkRxf0Z5NCtkAkgQZPIb4,629
@@ -82,7 +82,7 @@ gllm_inference/schema/lm_output.pyi,sha256=WP2LQrY0D03OJtFoaW_dGoJ_-yFUh2HbVlllg
 gllm_inference/schema/message.pyi,sha256=jJV6A0ihEcun2OhzyMtNkiHnf7d6v5R-GdpTBGfJ0AQ,2272
 gllm_inference/schema/model_id.pyi,sha256=3prO19l-FCSecRupe93ruXe91-Xw3GJOpbuQ66bijo0,5368
 gllm_inference/schema/reasoning.pyi,sha256=jbPxkDRHt0Vt-zdcc8lTT1l2hIE1Jm3HIHeNd0hfXGo,577
-gllm_inference/schema/token_usage.pyi,sha256=Eevs8S-yXoM7kQkkzhXHEvORU8DMGzdQynAamqtIoX4,323
+gllm_inference/schema/token_usage.pyi,sha256=WJiGQyz5qatzBK2b-sABLCyTRLCBbAvxCRcqSJOzu-8,3025
 gllm_inference/schema/tool_call.pyi,sha256=OWT9LUqs_xfUcOkPG0aokAAqzLYYDkfnjTa0zOWvugk,403
 gllm_inference/schema/tool_result.pyi,sha256=IJsU3n8y0Q9nFMEiq4RmLEIHueSiim0Oz_DlhKrTqto,287
 gllm_inference/schema/type_alias.pyi,sha256=qAljeBoeQEfT601maGe_mEpXD9inNzbGte1i6joQafc,740
@@ -90,8 +90,8 @@ gllm_inference/utils/__init__.pyi,sha256=RBTWDu1TDPpTd17fixcPYFv2L_vp4-IAOX0Isxg
 gllm_inference/utils/langchain.pyi,sha256=4AwFiVAO0ZpdgmqeC4Pb5NJwBt8vVr0MSUqLeCdTscc,1194
 gllm_inference/utils/validation.pyi,sha256=-RdMmb8afH7F7q4Ao7x6FbwaDfxUHn3hA3WiOgzB-3s,397
 gllm_inference.build/.gitignore,sha256=aEiIwOuxfzdCmLZe4oB1JsBmCUxwG8x-u-HBCV9JT8E,1
-gllm_inference.cp311-win_amd64.pyd,sha256=hYbjt2yhDh_GQb1f8NT0OAIm1xsWyZDWKt3XMvukpt8,2687488
-gllm_inference.pyi,sha256=HPIjsiX3usTXFeStYEslrHJDUtwpBBwl-Exzcf7y7Mc,3295
-gllm_inference_binary-0.5.7.dist-info/METADATA,sha256=5l9vCcKon7_P35hWo6n7VhczO9x6rBULOCosHG3l4sY,4531
-gllm_inference_binary-0.5.7.dist-info/WHEEL,sha256=-FZBVKyKauScY3vLa8vJR6hBCpAJfFykw2MOwlNKr1g,98
-gllm_inference_binary-0.5.7.dist-info/RECORD,,
+gllm_inference.cp311-win_amd64.pyd,sha256=ps5de1Zy0niIitf_qMVwZVsWtVk0NND-0Iysj9tDMxs,2732032
+gllm_inference.pyi,sha256=fsNCXXsB4E8WhP477yGq_QOJAfOyoZA4G2PfAMBav5Y,3324
+gllm_inference_binary-0.5.10b12.dist-info/METADATA,sha256=4VxkEu42rcmnLJChwVgqcxxC1PQutoY-Ba4X2352AoA,4535
+gllm_inference_binary-0.5.10b12.dist-info/WHEEL,sha256=-FZBVKyKauScY3vLa8vJR6hBCpAJfFykw2MOwlNKr1g,98
+gllm_inference_binary-0.5.10b12.dist-info/RECORD,,