gllm-inference-binary 0.5.57__cp313-cp313-manylinux_2_31_x86_64.whl → 0.5.65__cp313-cp313-manylinux_2_31_x86_64.whl

gllm_inference/builder/build_lm_invoker.pyi

@@ -1,4 +1,4 @@
-from gllm_inference.lm_invoker import AnthropicLMInvoker as AnthropicLMInvoker, AzureOpenAILMInvoker as AzureOpenAILMInvoker, BedrockLMInvoker as BedrockLMInvoker, DatasaurLMInvoker as DatasaurLMInvoker, GoogleLMInvoker as GoogleLMInvoker, LangChainLMInvoker as LangChainLMInvoker, LiteLLMLMInvoker as LiteLLMLMInvoker, OpenAIChatCompletionsLMInvoker as OpenAIChatCompletionsLMInvoker, OpenAICompatibleLMInvoker as OpenAICompatibleLMInvoker, OpenAILMInvoker as OpenAILMInvoker, PortkeyLMInvoker as PortkeyLMInvoker, XAILMInvoker as XAILMInvoker
+from gllm_inference.lm_invoker import AnthropicLMInvoker as AnthropicLMInvoker, AzureOpenAILMInvoker as AzureOpenAILMInvoker, BedrockLMInvoker as BedrockLMInvoker, DatasaurLMInvoker as DatasaurLMInvoker, GoogleLMInvoker as GoogleLMInvoker, LangChainLMInvoker as LangChainLMInvoker, LiteLLMLMInvoker as LiteLLMLMInvoker, OpenAIChatCompletionsLMInvoker as OpenAIChatCompletionsLMInvoker, OpenAICompatibleLMInvoker as OpenAICompatibleLMInvoker, OpenAILMInvoker as OpenAILMInvoker, PortkeyLMInvoker as PortkeyLMInvoker, SeaLionLMInvoker as SeaLionLMInvoker, XAILMInvoker as XAILMInvoker
 from gllm_inference.lm_invoker.lm_invoker import BaseLMInvoker as BaseLMInvoker
 from gllm_inference.schema.model_id import ModelId as ModelId, ModelProvider as ModelProvider
 from typing import Any
@@ -119,6 +119,15 @@ def build_lm_invoker(model_id: str | ModelId, credentials: str | dict[str, Any]
         ```
         The credentials can also be provided through the `AZURE_OPENAI_API_KEY` environment variable.
 
+        # Using SEA-LION
+        ```python
+        lm_invoker = build_lm_invoker(
+            model_id="sea-lion/aisingapore/Qwen-SEA-LION-v4-32B-IT",
+            credentials="sk-..."
+        )
+        ```
+        The credentials can also be provided through the `SEA_LION_API_KEY` environment variable.
+
         # Using LangChain
         ```python
         lm_invoker = build_lm_invoker(

gllm_inference/lm_invoker/__init__.pyi

@@ -9,6 +9,7 @@ from gllm_inference.lm_invoker.openai_chat_completions_lm_invoker import OpenAIC
 from gllm_inference.lm_invoker.openai_compatible_lm_invoker import OpenAICompatibleLMInvoker as OpenAICompatibleLMInvoker
 from gllm_inference.lm_invoker.openai_lm_invoker import OpenAILMInvoker as OpenAILMInvoker
 from gllm_inference.lm_invoker.portkey_lm_invoker import PortkeyLMInvoker as PortkeyLMInvoker
+from gllm_inference.lm_invoker.sea_lion_lm_invoker import SeaLionLMInvoker as SeaLionLMInvoker
 from gllm_inference.lm_invoker.xai_lm_invoker import XAILMInvoker as XAILMInvoker
 
-__all__ = ['AnthropicLMInvoker', 'AzureOpenAILMInvoker', 'BedrockLMInvoker', 'DatasaurLMInvoker', 'GoogleLMInvoker', 'LangChainLMInvoker', 'LiteLLMLMInvoker', 'OpenAIChatCompletionsLMInvoker', 'OpenAICompatibleLMInvoker', 'OpenAILMInvoker', 'PortkeyLMInvoker', 'XAILMInvoker']
+__all__ = ['AnthropicLMInvoker', 'AzureOpenAILMInvoker', 'BedrockLMInvoker', 'DatasaurLMInvoker', 'GoogleLMInvoker', 'LangChainLMInvoker', 'LiteLLMLMInvoker', 'OpenAIChatCompletionsLMInvoker', 'OpenAICompatibleLMInvoker', 'OpenAILMInvoker', 'PortkeyLMInvoker', 'SeaLionLMInvoker', 'XAILMInvoker']

gllm_inference/lm_invoker/anthropic_lm_invoker.pyi

@@ -1,12 +1,14 @@
 from _typeshed import Incomplete
+from anthropic.types import ContentBlockStopEvent as ContentBlockStopEvent, Message as Message, RawContentBlockDeltaEvent as RawContentBlockDeltaEvent, RawContentBlockStartEvent as RawContentBlockStartEvent
 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, BatchStatus as BatchStatus, LMInput as LMInput, LMOutput as LMOutput, Message as Message, ModelId as ModelId, ModelProvider as ModelProvider, Reasoning as Reasoning, ResponseSchema as ResponseSchema, ThinkingEvent as ThinkingEvent, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
+from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, BatchStatus as BatchStatus, LMInput as LMInput, LMOutput as LMOutput, ModelId as ModelId, ModelProvider as ModelProvider, Reasoning as Reasoning, ResponseSchema as ResponseSchema, ThinkingEvent as ThinkingEvent, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
 from langchain_core.tools import Tool as LangChainTool
+from pydantic import BaseModel as BaseModel
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete

gllm_inference/lm_invoker/batch/batch_operations.pyi

@@ -104,11 +104,12 @@ class BatchOperations:
         Returns:
             BatchStatus: The status of the batch job.
         """
-    async def retrieve(self, batch_id: str) -> dict[str, LMOutput]:
+    async def retrieve(self, batch_id: str, **kwargs: Any) -> dict[str, LMOutput]:
         """Retrieves the results of a batch job.
 
         Args:
             batch_id (str): The ID of the batch job to get the results of.
+            **kwargs (Any): Additional keyword arguments.
 
         Returns:
             dict[str, LMOutput]: The results of the batch job.

gllm_inference/lm_invoker/google_lm_invoker.pyi

@@ -6,8 +6,9 @@ from gllm_inference.constants import GOOGLE_SCOPES as GOOGLE_SCOPES, SECONDS_TO_
 from gllm_inference.exceptions import BaseInvokerError as BaseInvokerError, convert_http_status_to_base_invoker_error as convert_http_status_to_base_invoker_error
 from gllm_inference.exceptions.provider_error_map import GOOGLE_ERROR_MAPPING as GOOGLE_ERROR_MAPPING
 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, LMOutput as LMOutput, Message as Message, MessageRole as MessageRole, ModelId as ModelId, ModelProvider as ModelProvider, Reasoning as Reasoning, ResponseSchema as ResponseSchema, ThinkingEvent as ThinkingEvent, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
+from gllm_inference.lm_invoker.schema.google import InputType as InputType, JobState as JobState, Key as Key
+from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, BatchStatus as BatchStatus, LMInput as LMInput, LMOutput as LMOutput, Message as Message, MessageRole as MessageRole, ModelId as ModelId, ModelProvider as ModelProvider, Reasoning as Reasoning, ResponseSchema as ResponseSchema, ThinkingEvent as ThinkingEvent, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
+from google.genai.types import GenerateContentResponse as GenerateContentResponse
 from langchain_core.tools import Tool as LangChainTool
 from typing import Any
 
@@ -15,7 +16,12 @@ SUPPORTED_ATTACHMENTS: Incomplete
 DEFAULT_THINKING_BUDGET: int
 REQUIRE_THINKING_MODEL_PREFIX: Incomplete
 IMAGE_GENERATION_MODELS: Incomplete
-YOUTUBE_URL_PATTERN: Incomplete
+BATCH_STATUS_MAP: Incomplete
+
+class URLPattern:
+    """Defines specific Google related URL patterns."""
+    GOOGLE_FILE: Incomplete
+    YOUTUBE: Incomplete
 
 class GoogleLMInvoker(BaseLMInvoker):
     '''A language model invoker to interact with Google language models.
@@ -260,6 +266,94 @@ class GoogleLMInvoker(BaseLMInvoker):
         ```python
         lm_invoker = GoogleLMInvoker(..., retry_config=retry_config)
         ```
+
+    Batch processing:
+        The `GoogleLMInvoker` supports batch processing, which allows the language model to process multiple
+        requests in a single call. Batch processing is supported through the `batch` attribute.
+
+        Due to Google SDK limitations with batch processing:
+        1. Only inline requests are currently supported (not file-based or BigQuery sources).
+        2. The total size of all requests must be under 20MB.
+        3. Original request indices are not preserved in the results. The results are keyed by request index in the
+        format \'1\', \'2\', etc, in which order are preserved based on the original request order. If you want to use
+        custom request IDs, you can pass them as a list of strings to the `custom_request_ids` keyword argument
+
+        Usage example:
+        ```python
+        requests = {"1": "What color is the sky?", "2": "What color is the grass?"}
+        results = await lm_invoker.batch.invoke(requests)
+        ```
+
+        Output example:
+        ```python
+        {
+            "1": LMOutput(outputs=[LMOutputItem(type="text", output="The sky is blue.")]),
+            "2": LMOutput(finish_details={"type": "error", "message": "..."}),
+        }
+        ```
+
+        The `GoogleLMInvoker` also supports the following standalone batch processing operations:
+
+        1. Create a batch job:
+            ```python
+            requests = {"1": "What color is the sky?", "2": "What color is the grass?"}
+            batch_id = await lm_invoker.batch.create(requests)
+            ```
+
+        2. Get the status of a batch job:
+            ```python
+            status = await lm_invoker.batch.status(batch_id)
+            ```
+
+        3. Retrieve the results of a batch job:
+
+            In default, the results will be keyed by request index in the format \'1\', \'2\', etc,
+            in which order are preserved based on the original request order.
+
+
+            ```python
+            results = await lm_invoker.batch.retrieve(batch_id)
+            ```
+
+            Output example:
+            ```python
+            {
+                "1": LMOutput(outputs=[LMOutputItem(type="text", output="The sky is blue.")]),
+                "2": LMOutput(finish_details={"type": "error", "error": {"message": "...", ...}, ...}),
+            }
+            ```
+
+            If you pass custom_request_ids to the create method, the results will be keyed by the custom_request_ids.
+            ```python
+            results = await lm_invoker.batch.retrieve(batch_id, custom_request_ids=["request_1", "request_2"])
+            ```
+
+            Output example:
+            ```python
+            {
+                "request_1": LMOutput(outputs=[LMOutputItem(type="text", output="The sky is blue.")]),
+                "request_2": LMOutput(finish_details={"type": "error", "error": {"message": "...", ...}, ...}),
+            }
+            ```
+
+        4. List the batch jobs:
+            ```python
+            batch_jobs = await lm_invoker.batch.list()
+            ```
+
+            Output example:
+            ```python
+            [
+                {"id": "batch_123", "status": "finished"},
+                {"id": "batch_456", "status": "in_progress"},
+                {"id": "batch_789", "status": "canceling"},
+            ]
+            ```
+
+        5. Cancel a batch job:
+            ```python
+            await lm_invoker.batch.cancel(batch_id)
+            ```
     '''
     client_params: Incomplete
     image_generation: Incomplete

gllm_inference/lm_invoker/lm_invoker.pyi

@@ -133,6 +133,24 @@ class BaseLMInvoker(ABC, metaclass=abc.ABCMeta):
         This method clears the response schema for the language model by calling the `set_response_schema` method with
         None.
         """
+    async def count_input_tokens(self, messages: LMInput) -> int:
+        """Counts the input tokens for an invocation request inputs.
+
+        This method counts the input tokens for an invocation request inputs. This method is useful for:
+        1. Estimating the cost of an invocation request before invoking the language model.
+        2. Checking if the invocation request is too large to be processed by the language model.
+
+        Args:
+            messages (LMInput): The input messages for the language model.
+                1. If a list of Message objects is provided, it is used as is.
+                2. If a list of MessageContent or a string is provided, it is converted into a user message.
+
+        Returns:
+            int: The number of input tokens for the invocation request.
+
+        Raises:
+            TimeoutError: If the invocation times out.
+        """
     async def invoke(self, messages: LMInput, hyperparameters: dict[str, Any] | None = None, event_emitter: EventEmitter | None = None) -> str | LMOutput:
         """Invokes the language model.
 

gllm_inference/lm_invoker/openai_lm_invoker.pyi

@@ -7,6 +7,9 @@ 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 ActivityEvent as ActivityEvent, Attachment as Attachment, AttachmentType as AttachmentType, CodeEvent as CodeEvent, CodeExecResult as CodeExecResult, LMOutput as LMOutput, MCPCall as MCPCall, MCPCallActivity as MCPCallActivity, MCPListToolsActivity as MCPListToolsActivity, MCPServer as MCPServer, Message as Message, MessageRole as MessageRole, ModelId as ModelId, ModelProvider as ModelProvider, Reasoning as Reasoning, ResponseSchema as ResponseSchema, ThinkingEvent as ThinkingEvent, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult, WebSearchActivity as WebSearchActivity
 from langchain_core.tools import Tool as LangChainTool
+from openai import AsyncStream as AsyncStream
+from openai.types.responses import Response as Response, ResponseFunctionWebSearch as ResponseFunctionWebSearch, ResponseOutputItem as ResponseOutputItem
+from openai.types.responses.response_output_item import McpCall as McpCall, McpListTools as McpListTools
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete

gllm_inference/lm_invoker/schema/google.pyi

@@ -7,8 +7,10 @@ class Key:
     FUNCTION: str
     FUNCTION_CALL: str
     HTTP_OPTIONS: str
+    ID: str
     NAME: str
     RETRY_OPTIONS: str
+    STATUS: str
     SYSTEM_INSTRUCTION: str
     THINKING_CONFIG: str
     TIMEOUT: str
@@ -16,9 +18,19 @@ class Key:
     RESPONSE_SCHEMA: str
     RESPONSE_MIME_TYPE: str
     VERTEXAI: str
+    CUSTOM_REQUEST_IDS: str
 
 class InputType:
     """Defines valid input types in Google."""
     APPLICATION_JSON: str
     MODEL: str
     USER: str
+
+class JobState:
+    """Defines valid output types in Google."""
+    JOB_STATE_CANCELLED: str
+    JOB_STATE_EXPIRED: str
+    JOB_STATE_FAILED: str
+    JOB_STATE_PENDING: str
+    JOB_STATE_RUNNING: str
+    JOB_STATE_SUCCEEDED: str

gllm_inference/lm_invoker/schema/openai.pyi

@@ -80,6 +80,7 @@ class OutputType:
     FIND_IN_PAGE: str
     FUNCTION_CALL: str
     IMAGE: str
+    IMAGE_GENERATION_CALL: str
     INCOMPLETE: str
     ITEM_DONE: str
     MCP_CALL: str

gllm_inference/lm_invoker/sea_lion_lm_invoker.pyi

@@ -0,0 +1,48 @@
+from _typeshed import Incomplete
+from gllm_core.schema.tool import Tool as Tool
+from gllm_core.utils import RetryConfig as RetryConfig
+from gllm_inference.constants import INVOKER_PROPAGATED_MAX_RETRIES as INVOKER_PROPAGATED_MAX_RETRIES
+from gllm_inference.lm_invoker.openai_chat_completions_lm_invoker import OpenAIChatCompletionsLMInvoker as OpenAIChatCompletionsLMInvoker
+from gllm_inference.lm_invoker.schema.openai_chat_completions 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 LangChainTool
+from typing import Any
+
+SEA_LION_URL: str
+SUPPORTED_ATTACHMENTS: Incomplete
+
+class SeaLionLMInvoker(OpenAIChatCompletionsLMInvoker):
+    """A language model invoker to interact with SEA-LION API.
+
+    Attributes:
+        model_id (str): The model ID of the language model.
+        model_provider (str): The provider of the language model.
+        model_name (str): The name of the language model.
+        client_kwargs (dict[str, Any]): The keyword arguments for the OpenAI client.
+        default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the model.
+        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.
+        retry_config (RetryConfig | None): The retry configuration for the language model.
+    """
+    client_kwargs: 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 | LangChainTool] | None = None, response_schema: ResponseSchema | None = None, output_analytics: bool = False, retry_config: RetryConfig | None = None) -> None:
+        """Initializes a new instance of the SeaLionLMInvoker class.
+
+        Args:
+            model_name (str): The name of the SEA-LION language model.
+            api_key (str | None, optional): The API key for authenticating with the SEA-LION API.
+                Defaults to None, in which case the `SEA_LION_API_KEY` environment variable will be used.
+            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 | 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.
+            output_analytics (bool, optional): Whether to output the invocation analytics. Defaults to False.
+            retry_config (RetryConfig | None, optional): The retry configuration for the language model.
+                Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+        """

gllm_inference/model/__init__.pyi

@@ -7,6 +7,7 @@ from gllm_inference.model.em.voyage_em import VoyageEM as VoyageEM
 from gllm_inference.model.lm.anthropic_lm import AnthropicLM as AnthropicLM
 from gllm_inference.model.lm.google_lm import GoogleLM as GoogleLM
 from gllm_inference.model.lm.openai_lm import OpenAILM as OpenAILM
+from gllm_inference.model.lm.sea_lion_lm import SeaLionLM as SeaLionLM
 from gllm_inference.model.lm.xai_lm import XAILM as XAILM
 
-__all__ = ['AnthropicLM', 'CohereEM', 'GoogleEM', 'GoogleLM', 'JinaEM', 'OpenAIEM', 'OpenAILM', 'TwelveLabsEM', 'VoyageEM', 'XAILM']
+__all__ = ['AnthropicLM', 'CohereEM', 'GoogleEM', 'GoogleLM', 'JinaEM', 'OpenAIEM', 'OpenAILM', 'SeaLionLM', 'TwelveLabsEM', 'VoyageEM', 'XAILM']

gllm_inference/model/lm/sea_lion_lm.pyi

@@ -0,0 +1,16 @@
+class SeaLionLM:
+    '''Defines SEA-LION language model names constants.
+
+    Usage example:
+    ```python
+    from gllm_inference.model import SeaLionLM
+    from gllm_inference.lm_invoker import SeaLionLMInvoker
+
+    lm_invoker = SeaLionLMInvoker(SeaLionLM.GEMMA_SEA_LION_V4_27B_IT)
+    response = await lm_invoker.invoke("Hello, world!")
+    ```
+    '''
+    GEMMA_SEA_LION_V4_27B_IT: str
+    LLAMA_SEA_LION_V3_5_70B_R: str
+    LLAMA_SEA_LION_V3_70B_IT: str
+    QWEN_SEA_LION_V4_32B_IT: str

gllm_inference/prompt_builder/prompt_builder.pyi

@@ -1,6 +1,6 @@
 from _typeshed import Incomplete
 from gllm_inference.prompt_builder.format_strategy import JinjaFormatStrategy as JinjaFormatStrategy, StringFormatStrategy as StringFormatStrategy
-from gllm_inference.schema import JinjaEnvType as JinjaEnvType, Message as Message, MessageContent as MessageContent, MessageRole as MessageRole
+from gllm_inference.schema import HistoryFormatter as HistoryFormatter, JinjaEnvType as JinjaEnvType, Message as Message, MessageContent as MessageContent, MessageRole as MessageRole
 from jinja2.sandbox import SandboxedEnvironment as SandboxedEnvironment
 from typing import Any
 
@@ -13,13 +13,15 @@ class PromptBuilder:
         prompt_key_set (set[str]): A set of expected keys that must be present in the prompt templates.
         key_defaults (dict[str, str]): Default values for the keys in the prompt templates.
         strategy (BasePromptFormattingStrategy): The format strategy to be used for formatting the prompt.
+        history_formatter (HistoryFormatter): The history formatter to be used for formatting the history.
     """
     key_defaults: Incomplete
     system_template: Incomplete
     user_template: Incomplete
+    history_formatter: Incomplete
     strategy: Incomplete
     prompt_key_set: Incomplete
-    def __init__(self, system_template: str = '', user_template: str = '', key_defaults: dict[str, str] | None = None, ignore_extra_keys: bool | None = None, use_jinja: bool = False, jinja_env: JinjaEnvType | SandboxedEnvironment = ...) -> None:
+    def __init__(self, system_template: str = '', user_template: str = '', key_defaults: dict[str, str] | None = None, ignore_extra_keys: bool | None = None, history_formatter: HistoryFormatter | None = None, use_jinja: bool | None = False, jinja_env: JinjaEnvType | SandboxedEnvironment | None = None) -> None:
         """Initializes a new instance of the PromptBuilder class.
 
         Args:
@@ -32,6 +34,8 @@ class PromptBuilder:
                 Defaults to None, in which case no default values will be assigned to the keys.
             ignore_extra_keys (bool | None, optional): Deprecated parameter. Will be removed in v0.6. Extra keys
                 will always raise a warning only instead of raising an error.
+            history_formatter (HistoryFormatter | None, optional): The history formatter to be used for formatting
+                the history. Defaults to None, in which case the history will be used as is.
             use_jinja (bool, optional): Whether to use Jinja for rendering the prompt templates.
                 Defaults to False.
             jinja_env (JinjaEnvType | SandboxedEnvironment, optional): The environment for Jinja rendering.

gllm_inference/schema/__init__.pyi

@@ -4,6 +4,7 @@ from gllm_inference.schema.code_exec_result import CodeExecResult as CodeExecRes
 from gllm_inference.schema.config import TruncationConfig as TruncationConfig
 from gllm_inference.schema.enums import AttachmentType as AttachmentType, BatchStatus as BatchStatus, EmitDataType as EmitDataType, JinjaEnvType as JinjaEnvType, LMEventType as LMEventType, LMEventTypeSuffix as LMEventTypeSuffix, LMOutputType as LMOutputType, MessageRole as MessageRole, TruncateSide as TruncateSide
 from gllm_inference.schema.events import ActivityEvent as ActivityEvent, CodeEvent as CodeEvent, ThinkingEvent as ThinkingEvent
+from gllm_inference.schema.formatter import HistoryFormatter as HistoryFormatter
 from gllm_inference.schema.lm_input import LMInput as LMInput
 from gllm_inference.schema.lm_output import LMOutput as LMOutput, LMOutputData as LMOutputData, LMOutputItem as LMOutputItem
 from gllm_inference.schema.mcp import MCPCall as MCPCall, MCPServer as MCPServer
@@ -15,4 +16,4 @@ 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, MessageContent as MessageContent, ResponseSchema as ResponseSchema, Vector as Vector
 
-__all__ = ['Activity', 'ActivityEvent', 'Attachment', 'AttachmentType', 'BatchStatus', 'CodeEvent', 'CodeExecResult', 'EMContent', 'EmitDataType', 'InputTokenDetails', 'JinjaEnvType', 'LMEventType', 'LMEventTypeSuffix', 'LMInput', 'LMOutput', 'LMOutputItem', 'LMOutputData', 'LMOutputType', 'MCPCall', 'MCPCallActivity', 'MCPListToolsActivity', 'MCPServer', 'Message', 'MessageContent', 'MessageRole', 'ModelId', 'ModelProvider', 'OutputTokenDetails', 'Reasoning', 'ResponseSchema', 'ThinkingEvent', 'TokenUsage', 'ToolCall', 'ToolResult', 'TruncateSide', 'TruncationConfig', 'Vector', 'WebSearchActivity']
+__all__ = ['Activity', 'ActivityEvent', 'Attachment', 'AttachmentType', 'BatchStatus', 'CodeEvent', 'CodeExecResult', 'EMContent', 'EmitDataType', 'HistoryFormatter', 'InputTokenDetails', 'JinjaEnvType', 'LMEventType', 'LMEventTypeSuffix', 'LMInput', 'LMOutput', 'LMOutputItem', 'LMOutputData', 'LMOutputType', 'MCPCall', 'MCPCallActivity', 'MCPListToolsActivity', 'MCPServer', 'Message', 'MessageContent', 'MessageRole', 'ModelId', 'ModelProvider', 'OutputTokenDetails', 'Reasoning', 'ResponseSchema', 'ThinkingEvent', 'TokenUsage', 'ToolCall', 'ToolResult', 'TruncateSide', 'TruncationConfig', 'Vector', 'WebSearchActivity']

gllm_inference/schema/attachment.pyi

@@ -1,7 +1,9 @@
 from _typeshed import Incomplete
 from pydantic import BaseModel
+from typing import Any
 
 HEX_REPR_LENGTH: int
+METADATA_ITEM_REPR_LENGTH: int
 logger: Incomplete
 
 class Attachment(BaseModel):
@@ -13,68 +15,80 @@ class Attachment(BaseModel):
         mime_type (str): The mime type of the file attachment.
         extension (str): The extension of the file attachment.
         url (str | None): The URL of the file attachment. Defaults to None.
+        metadata (dict[str, Any]): The metadata of the file attachment. Defaults to an empty dictionary.
     """
     data: bytes
     filename: str
     mime_type: str
     extension: str
     url: str | None
+    metadata: dict[str, Any]
     @classmethod
-    def from_bytes(cls, bytes: bytes, filename: str | None = None) -> Attachment:
+    def from_bytes(cls, bytes: bytes, filename: str | None = None, metadata: dict[str, Any] | None = None) -> Attachment:
         """Creates an Attachment from bytes.
 
         Args:
             bytes (bytes): The bytes of the file.
             filename (str | None, optional): The filename of the file. Defaults to None,
                 in which case the filename will be derived from the extension.
+            metadata (dict[str, Any] | None, optional): The metadata of the file attachment. Defaults to None,
+                in which case an empty dictionary will be used.
 
         Returns:
             Attachment: The instantiated Attachment.
         """
     @classmethod
-    def from_base64(cls, base64_data: str, filename: str | None = None) -> Attachment:
+    def from_base64(cls, base64_data: str, filename: str | None = None, metadata: dict[str, Any] | None = None) -> Attachment:
         """Creates an Attachment from a base64 string.
 
         Args:
             base64_data (str): The base64 string of the file.
             filename (str | None, optional): The filename of the file. Defaults to None,
                 in which case the filename will be derived from the mime type.
+            metadata (dict[str, Any] | None, optional): The metadata of the file attachment. Defaults to None,
+                in which case an empty dictionary will be used.
 
         Returns:
             Attachment: The instantiated Attachment.
         """
     @classmethod
-    def from_data_url(cls, data_url: str, filename: str | None = None) -> Attachment:
+    def from_data_url(cls, data_url: str, filename: str | None = None, metadata: dict[str, Any] | None = None) -> Attachment:
         """Creates an Attachment from a data URL (data:[mime/type];base64,[bytes]).
 
         Args:
             data_url (str): The data URL of the file.
             filename (str | None, optional): The filename of the file. Defaults to None,
                 in which case the filename will be derived from the mime type.
+            metadata (dict[str, Any] | None, optional): The metadata of the file attachment. Defaults to None,
+                in which case an empty dictionary will be used.
 
         Returns:
             Attachment: The instantiated Attachment.
         """
     @classmethod
-    def from_url(cls, url: str, filename: str | None = None) -> Attachment:
+    def from_url(cls, url: str, filename: str | None = None, metadata: dict[str, Any] | None = None) -> Attachment:
         """Creates an Attachment from a URL.
 
         Args:
             url (str): The URL of the file.
             filename (str | None, optional): The filename of the file. Defaults to None,
                 in which case the filename will be derived from the URL.
+            metadata (dict[str, Any] | None, optional): The metadata of the file attachment. Defaults to None,
+                in which case an empty dictionary will be used.
 
         Returns:
             Attachment: The instantiated Attachment.
         """
     @classmethod
-    def from_path(cls, path: str, filename: str | None = None) -> Attachment:
+    def from_path(cls, path: str, filename: str | None = None, metadata: dict[str, Any] | None = None) -> Attachment:
         """Creates an Attachment from a path.
 
         Args:
             path (str): The path to the file.
             filename (str | None, optional): The filename of the file. Defaults to None,
                 in which case the filename will be derived from the path.
+            metadata (dict[str, Any] | None, optional): The metadata of the file attachment. Defaults to None,
+                in which case an empty dictionary will be used.
 
         Returns:
             Attachment: The instantiated Attachment.

gllm_inference/schema/enums.pyi

@@ -9,9 +9,11 @@ class AttachmentType(StrEnum):
 
 class BatchStatus(StrEnum):
     """Defines the status of a batch job."""
-    CANCELING = 'canceling'
     IN_PROGRESS = 'in_progress'
     FINISHED = 'finished'
+    FAILED = 'failed'
+    CANCELING = 'canceling'
+    EXPIRED = 'expired'
     UNKNOWN = 'unknown'
 
 class LMEventType(StrEnum):

gllm_inference/schema/formatter.pyi

@@ -0,0 +1,31 @@
+from gllm_inference.schema.enums import MessageRole as MessageRole
+from gllm_inference.schema.message import Message as Message
+from gllm_inference.schema.type_alias import MessageContent as MessageContent
+from pydantic import BaseModel
+
+class HistoryFormatter(BaseModel):
+    """Configuration for history formatting.
+
+    Attributes:
+        prefix_user_message (str): Prefix for user messages.
+        suffix_user_message (str): Suffix for user messages.
+        prefix_assistant_message (str): Prefix for assistant messages.
+        suffix_assistant_message (str): Suffix for assistant messages.
+    """
+    prefix_user_message: str
+    suffix_user_message: str
+    prefix_assistant_message: str
+    suffix_assistant_message: str
+    def format_history(self, history: list[Message]) -> list[Message]:
+        """Formats a list of messages based on their roles.
+
+        This method formats each message in the history list by applying the appropriate
+        formatting based on the message role (user or assistant). Other message types
+        are added to the result without modification.
+
+        Args:
+            history (list[Message]): The list of messages to format.
+
+        Returns:
+            list[Message]: A new list containing the formatted messages.
+        """

gllm_inference/schema/model_id.pyi

@@ -21,6 +21,7 @@ class ModelProvider(StrEnum):
     PORTKEY = 'portkey'
     OPENAI_CHAT_COMPLETIONS = 'openai-chat-completions'
     OPENAI_COMPATIBLE = 'openai-compatible'
+    SEA_LION = 'sea-lion'
     TWELVELABS = 'twelvelabs'
     VOYAGE = 'voyage'
     XAI = 'xai'

gllm_inference.pyi

@@ -33,6 +33,7 @@ import gllm_inference.lm_invoker.OpenAIChatCompletionsLMInvoker
 import gllm_inference.lm_invoker.OpenAICompatibleLMInvoker
 import gllm_inference.lm_invoker.OpenAILMInvoker
 import gllm_inference.lm_invoker.PortkeyLMInvoker
+import gllm_inference.lm_invoker.SeaLionLMInvoker
 import gllm_inference.lm_invoker.XAILMInvoker
 import gllm_inference.prompt_builder.PromptBuilder
 import gllm_inference.output_parser.JSONOutputParser
@@ -82,6 +83,7 @@ import voyageai
 import voyageai.client_async
 import http
 import http.HTTPStatus
+import __future__
 import uuid
 import gllm_core.constants
 import gllm_core.event
@@ -113,7 +115,6 @@ import time
 import jsonschema
 import gllm_inference.lm_invoker.batch.BatchOperations
 import gllm_inference.schema.MessageContent
-import __future__
 import gllm_inference.schema.ActivityEvent
 import gllm_inference.schema.CodeEvent
 import gllm_inference.schema.CodeExecResult
@@ -135,6 +136,7 @@ import jinja2.sandbox
 import gllm_inference.schema.JinjaEnvType
 import gllm_inference.prompt_builder.format_strategy.JinjaFormatStrategy
 import gllm_inference.prompt_builder.format_strategy.StringFormatStrategy
+import gllm_inference.schema.HistoryFormatter
 import transformers
 import gllm_inference.prompt_formatter.HuggingFacePromptFormatter
 import traceback
@@ -148,6 +150,7 @@ import pathlib.Path
 import filetype
 import magic
 import requests
+import gllm_core.schema.chunk
 import binascii
 import fnmatch
 import importlib

{gllm_inference_binary-0.5.57.dist-info → gllm_inference_binary-0.5.65.dist-info}/METADATA

@@ -1,8 +1,8 @@
 Metadata-Version: 2.2
 Name: gllm-inference-binary
-Version: 0.5.57
+Version: 0.5.65
 Summary: A library containing components related to model inferences in Gen AI applications.
-Author-email: Henry Wicaksono <henry.wicaksono@gdplabs.id>, Resti Febrina <resti.febrina@gdplabs.id>
+Author-email: Henry Wicaksono <henry.wicaksono@gdplabs.id>, "Delfia N. A. Putri" <delfia.n.a.putri@gdplabs.id>
 Requires-Python: <3.14,>=3.11
 Description-Content-Type: text/markdown
 Requires-Dist: poetry<3.0.0,>=2.1.3
@@ -33,7 +33,7 @@ Requires-Dist: aioboto3<16.0.0,>=15.0.0; extra == "bedrock"
 Provides-Extra: cohere
 Requires-Dist: cohere<6.0.0,>=5.18.0; extra == "cohere"
 Provides-Extra: datasaur
-Requires-Dist: openai<2.0.0,>=1.98.0; extra == "datasaur"
+Requires-Dist: openai<3.0.0,>=2.7.0; extra == "datasaur"
 Provides-Extra: google
 Requires-Dist: google-genai<=1.36,>=1.23; extra == "google"
 Provides-Extra: huggingface
@@ -42,7 +42,7 @@ Requires-Dist: transformers<5.0.0,>=4.52.0; extra == "huggingface"
 Provides-Extra: litellm
 Requires-Dist: litellm<2.0.0,>=1.69.2; extra == "litellm"
 Provides-Extra: openai
-Requires-Dist: openai<2.0.0,>=1.98.0; extra == "openai"
+Requires-Dist: openai<3.0.0,>=2.7.0; extra == "openai"
 Provides-Extra: portkey-ai
 Requires-Dist: portkey-ai<2.0.0,>=1.14.4; extra == "portkey-ai"
 Provides-Extra: twelvelabs

{gllm_inference_binary-0.5.57.dist-info → gllm_inference_binary-0.5.65.dist-info}/RECORD

@@ -1,11 +1,11 @@
-gllm_inference.cpython-313-x86_64-linux-gnu.so,sha256=PTkaVkJSAvooZuKI01fjrFMNkBqGNfVJIhnjB-eTk2U,5708888
-gllm_inference.pyi,sha256=U5ETTUzO_5DKsg4Zp7WQ-rnG_RJKCAqKaf6FoNNw2m4,5067
+gllm_inference.cpython-313-x86_64-linux-gnu.so,sha256=aMQyK0sMg5NfAmDJGrrX58UWy28FZQApYc78D87pK6k,5819608
+gllm_inference.pyi,sha256=3UoxsyZnC_6PhNzIKIp37rL6CNEw-Ve0RJHiG9jLOBY,5193
 gllm_inference/__init__.pyi,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 gllm_inference/constants.pyi,sha256=tBFhwE1at2gXMJ1bBM32eVIRgCJlB1uzg7ItXGx3RQE,316
 gllm_inference/builder/__init__.pyi,sha256=usz2lvfwO4Yk-ZGKXbCWG1cEr3nlQXxMNDNC-2yc1NM,500
 gllm_inference/builder/_build_invoker.pyi,sha256=v__-YT0jGmLqEsgl7Abk9we-wYWcyVFvlRN_Uu4vVak,848
 gllm_inference/builder/build_em_invoker.pyi,sha256=0IVcRGaciPBAFcgvGRC4-Kje_PPXOSug77LqxRc-x_U,5749
-gllm_inference/builder/build_lm_invoker.pyi,sha256=TBAr7Sk9Jgckdlfj69pYOzZFMsooL-FpfbPYt_kCcXU,8852
+gllm_inference/builder/build_lm_invoker.pyi,sha256=o4dIL90wDu94y_fIeUs2ei5Qkk1zZo5_nYBqL2EiHAE,9195
 gllm_inference/builder/build_lm_request_processor.pyi,sha256=KbQkcPa8C-yzyelht4mWLP8kDmh17itAT3tn8ZJB6pg,4144
 gllm_inference/builder/build_output_parser.pyi,sha256=_Lrq-bh1oPsb_Nwkkr_zyEUwIOMysRFZkvEtEM29LZM,936
 gllm_inference/catalog/__init__.pyi,sha256=JBkPGTyiiZ30GECzJBW-mW8LekWyY2qyzal3eW7ynaM,287
@@ -40,33 +40,34 @@ gllm_inference/exceptions/__init__.pyi,sha256=Upcuj7od2lkbdueQ0iMT2ktFYYi-KKTynT
 gllm_inference/exceptions/error_parser.pyi,sha256=IOfa--NpLUW5E9Qq0mwWi6ZpTAbUyyNe6iAqunBNGLI,1999
 gllm_inference/exceptions/exceptions.pyi,sha256=Bv996qLa_vju0Qjf4GewMxdkq8CV9LRZb0S6289DldA,5725
 gllm_inference/exceptions/provider_error_map.pyi,sha256=XPLWU42-r8MHZgg5ZkE80Gdqg3p8Z_JHvq_Na03iTqY,1243
-gllm_inference/lm_invoker/__init__.pyi,sha256=Ze9CxgGYguyz8BAU87_2JM-D4OZjlYAqktLI_B2tj_s,1488
-gllm_inference/lm_invoker/anthropic_lm_invoker.pyi,sha256=oU0dzg87OHTVPGhKBT8WdcdR0AzeJJNSSVvGKtY0UQU,15178
+gllm_inference/lm_invoker/__init__.pyi,sha256=dvwZQhO-7bHDOouXBXxujux9QQbu10ux-0vAsJI1DeI,1603
+gllm_inference/lm_invoker/anthropic_lm_invoker.pyi,sha256=_by_rLLdBpnpIyniBVCOEhjur5HzdBJ1UYoXdpxb8Lw,15409
 gllm_inference/lm_invoker/azure_openai_lm_invoker.pyi,sha256=drtMgbDzBQJrWXLiI2t5PNy7HtcW5Kuj0XR2b6rltjc,12936
 gllm_inference/lm_invoker/bedrock_lm_invoker.pyi,sha256=9Gz0U2c94UM9SOt-_e89_NqT_bDQ6wECRiJ9VTwsqfw,10739
 gllm_inference/lm_invoker/datasaur_lm_invoker.pyi,sha256=IqvDxBzwEf2z34FZcLKIH404y386Rnk5gsj2TcAm424,7878
-gllm_inference/lm_invoker/google_lm_invoker.pyi,sha256=wWnLHY-Q9Id6E5al4Wy57gIViTIJ-R-ZjLK1SFnESZw,16363
+gllm_inference/lm_invoker/google_lm_invoker.pyi,sha256=DrMIhGhWolSBH26jTkx9zaXTVcRBqxBwNWgzI3InODE,20063
 gllm_inference/lm_invoker/langchain_lm_invoker.pyi,sha256=OzISl89C2s-qB6VxNlMgf5dFRC-ooj30YCFfsZzcX4s,11887
 gllm_inference/lm_invoker/litellm_lm_invoker.pyi,sha256=0PZYitAljAjzyympqqNyN5fMyoakmqr1XIz1PE6NNc4,11176
-gllm_inference/lm_invoker/lm_invoker.pyi,sha256=pJ0-s37NqTHdFD7IijvNzJnQ7JXgrGxsEaXuS8cxz3s,8487
+gllm_inference/lm_invoker/lm_invoker.pyi,sha256=tgyv7A8K_36mYxL_Fkb7DFCZfJT_Hy08CKSiSJ88f0M,9370
 gllm_inference/lm_invoker/openai_chat_completions_lm_invoker.pyi,sha256=47lMyuzWKn57Ndt124-zzl9kKHa7IYzDu51iRdUPg6c,13636
 gllm_inference/lm_invoker/openai_compatible_lm_invoker.pyi,sha256=i5pMpZf4-r_7FQ1qfsqcjpc98sI-cPiqheuTfTEKxJs,4192
-gllm_inference/lm_invoker/openai_lm_invoker.pyi,sha256=OytpncpkOLdaSyFQ41GLh4_MioEbpsX0WiA8zc70Q6s,22313
+gllm_inference/lm_invoker/openai_lm_invoker.pyi,sha256=oKVs5UnJJxjWaO7Rd0mmwPfBqEFeZRI09mmRdONs3js,22618
 gllm_inference/lm_invoker/portkey_lm_invoker.pyi,sha256=ewxzRT-ekmvCFeb7ij840s4p18AO_LAKA-UP0ot12hs,14940
+gllm_inference/lm_invoker/sea_lion_lm_invoker.pyi,sha256=ElV7iKYWnI3J1CUYuHtvOTsJByMY_l2WF4Rc7IJsBjw,3485
 gllm_inference/lm_invoker/xai_lm_invoker.pyi,sha256=SVIsRGcqbRnR9sqoLYWwigoEumDib5m4cTaTJT98Uz4,12765
 gllm_inference/lm_invoker/batch/__init__.pyi,sha256=W4W-_yfk7lL20alREJai6GnwuQvdlKRfwQCX4mQK4XI,127
-gllm_inference/lm_invoker/batch/batch_operations.pyi,sha256=Oo7hoyPSfPZdy1mXvSdvtRndvq-XTIbPIjEoGvJj5C0,5372
+gllm_inference/lm_invoker/batch/batch_operations.pyi,sha256=cEdRVHAb5uiIPnlQudtaNzQ9kszI7cH6nD9mpMYQLco,5445
 gllm_inference/lm_invoker/schema/__init__.pyi,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 gllm_inference/lm_invoker/schema/anthropic.pyi,sha256=6lreMyHKRfZzX5NBYKnQf1Z6RzXBjTvqZj2VbMeaTLQ,1098
 gllm_inference/lm_invoker/schema/bedrock.pyi,sha256=FJLY-ZkkLUYDV48pfsLatnot4ev_xxz9xAayLK28CpU,1027
 gllm_inference/lm_invoker/schema/datasaur.pyi,sha256=WSuwOqL1j2ZioCZFC-gbB7vTRIZHQ3sU40c3ool5L6c,265
-gllm_inference/lm_invoker/schema/google.pyi,sha256=AIsNgq0ZZuicHmx4bL7z6q-946T05nWts3HUeA8hhHQ,505
+gllm_inference/lm_invoker/schema/google.pyi,sha256=MYjznjkKfNdh9XwTIrrK29tS4pkGPEU7WebVfCvFLGw,791
 gllm_inference/lm_invoker/schema/langchain.pyi,sha256=rZcIxuvABI4pKfyVvkRBRqfJJogZ67EFPydpubHt49c,429
-gllm_inference/lm_invoker/schema/openai.pyi,sha256=PS4QrNYa1NPq0s-3r8FI22a2Td0n70UONLxPn8rP_s4,2279
+gllm_inference/lm_invoker/schema/openai.pyi,sha256=y__q2_2ZcfqCkWBw3uA6iCEQKDCm7IJoLVo3GS-ck9c,2310
 gllm_inference/lm_invoker/schema/openai_chat_completions.pyi,sha256=8byBRZ4xyTidIQJsZqiSjp5t1X875Obe-aEbT0yYfuA,1199
 gllm_inference/lm_invoker/schema/portkey.pyi,sha256=NeRjHNd84HgE_ur2F3Cv6Jx30v6V7eQvI_iJiq4kuME,631
 gllm_inference/lm_invoker/schema/xai.pyi,sha256=cWnbJmDtllqRH3NXpQbiXgkNBcUXr8ksDSDywcgJebE,632
-gllm_inference/model/__init__.pyi,sha256=LTeBCSJJwCSd5Qrg7RZCXcp9fURNVNXFR5akk1ZZrTk,810
+gllm_inference/model/__init__.pyi,sha256=1gesoNUUHutpN2-DYmVrG5LZhpeSW96ciON3SjVEqqM,894
 gllm_inference/model/em/__init__.pyi,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 gllm_inference/model/em/cohere_em.pyi,sha256=fArRlV08NwbsJ_h6vpWr94XxUVBtbqW1Jh8s42LRXCo,488
 gllm_inference/model/em/google_em.pyi,sha256=ZPN5LmReO0bcTfnZixFooUTzgD-daNFPzfxzZ-5WzQQ,471
@@ -78,12 +79,13 @@ gllm_inference/model/lm/__init__.pyi,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZ
 gllm_inference/model/lm/anthropic_lm.pyi,sha256=dWfG-M_gD644yJ-LK_T8HnAT649j3Vx7TVof03XQimE,611
 gllm_inference/model/lm/google_lm.pyi,sha256=cMV5zYX8uwUF7pErv4pXnXD2G52umo3sxKwbSx7nFhQ,511
 gllm_inference/model/lm/openai_lm.pyi,sha256=u11zvvIS7-XaHKZ33cZxGQmT6cZ4DqK9Do8l7gFOUTc,618
+gllm_inference/model/lm/sea_lion_lm.pyi,sha256=k0xG5JGrecBDlNKWwNZb4BG0Ath_tGfy4fudr51-10w,492
 gllm_inference/model/lm/xai_lm.pyi,sha256=2ZEQ_--e_zsb23zZQ8bKdQShU7zChx5TrDKF8EpwEpU,506
 gllm_inference/output_parser/__init__.pyi,sha256=WQOOgsYnPk8vd-SOhFMMaVTzy4gkYrOAyT5gnAxv0A0,129
 gllm_inference/output_parser/json_output_parser.pyi,sha256=uulh91uQLMSb4ZXZhHYi9W9w7zGnmrOweEkL6wdDJN8,2933
 gllm_inference/output_parser/output_parser.pyi,sha256=Yzk7F26pH8Uc7FQZo4G6l67YkfppefUvaV9cNK-HyDs,948
 gllm_inference/prompt_builder/__init__.pyi,sha256=kshfBMvwIwiIvjxiGG5BrJZNvpPa8rhtkbHo5FPifBg,117
-gllm_inference/prompt_builder/prompt_builder.pyi,sha256=VQaw8nE8SPK1nvVVB4YjeqZSazigNPuWvOEcgGji0W8,4557
+gllm_inference/prompt_builder/prompt_builder.pyi,sha256=qJ6L81KMLdZvS2MmOGGjrJ9btmWwVPoRVl3NQxHq3fw,5010
 gllm_inference/prompt_builder/format_strategy/__init__.pyi,sha256=QhORHac3ySOPmL9k9kmCKL70vtaUtwkZEtGoRNWNuA8,308
 gllm_inference/prompt_builder/format_strategy/format_strategy.pyi,sha256=JSUl_7Ka08oDZPpslymkUa8pDzqNGIK2TlcVANspqrY,2273
 gllm_inference/prompt_builder/format_strategy/jinja_format_strategy.pyi,sha256=IAezLUiKSJMaoDyleo8pFnFqq8rBM_Q-lNXuAGvwXhI,2225
@@ -109,18 +111,19 @@ gllm_inference/realtime_chat/output_streamer/output_streamer.pyi,sha256=GPAw1wPS
 gllm_inference/request_processor/__init__.pyi,sha256=hVnfdNZnkTBJHnmLtN3Na4ANP0yK6AstWdIizVr2Apo,227
 gllm_inference/request_processor/lm_request_processor.pyi,sha256=VnYc8E3Iayyhw-rPnGPfTKuO3ohgFsS8HPrZJeyES5I,5889
 gllm_inference/request_processor/uses_lm_mixin.pyi,sha256=Yu0XPNuHxq1tWBviHTPw1oThojneFwGHepvGjBXxKQA,6382
-gllm_inference/schema/__init__.pyi,sha256=hgRrwTocQ8b5MDDosSQN8zEuarGckpVply1OwEbrd28,2404
+gllm_inference/schema/__init__.pyi,sha256=OOZE9H4DgUqOsk5_01hXb5opkAwBmp1TvhqcV2SrqJY,2505
 gllm_inference/schema/activity.pyi,sha256=JnO2hqj91P5Tc6qb4pbkEMrHer2u5owiCvhl-igcQKQ,2303
-gllm_inference/schema/attachment.pyi,sha256=EHV3uPB3lVZCMNxySVuIoB6ikAiY7THSNh77q3ZtkVY,3246
+gllm_inference/schema/attachment.pyi,sha256=oCopoxiPgGSkCRdPsqmjcMofTawfbdCDxaPdo6mits0,4509
 gllm_inference/schema/code_exec_result.pyi,sha256=ZTHh6JtRrPIdQ059P1UAiD2L-tAO1_S5YcMsAXfJ5A0,559
 gllm_inference/schema/config.pyi,sha256=rAL_UeXyQeXVk1P2kqd8vFWOMwmKenfpQLtvMP74t9s,674
-gllm_inference/schema/enums.pyi,sha256=jByrR0Y84-WZ3KDPUjuOyfecouUATyO-A8rdehKPjgs,2065
+gllm_inference/schema/enums.pyi,sha256=rLTlnhtdTIKT_Q8p-ukKpCMhRk9NbqcR_ylZ0-8UivA,2111
 gllm_inference/schema/events.pyi,sha256=XQEy5SqYoutq2DguwCCTGi5DHrlnDai6nJElAYsj1gk,4638
+gllm_inference/schema/formatter.pyi,sha256=qPQ1oSnmSgr7yBsBBMe-aehLLk9lKI9OPZ8Og-EbzdI,1281
 gllm_inference/schema/lm_input.pyi,sha256=A5pjz1id6tP9XRNhzQrbmzd66C_q3gzo0UP8rCemz6Q,193
 gllm_inference/schema/lm_output.pyi,sha256=ec5ErE68PVthNHhHlaVHsiAN3nn03yBqRR7HVabsf68,11401
 gllm_inference/schema/mcp.pyi,sha256=Vwu8E2BDl6FvvnI42gIyY3Oki1BdwRE3Uh3aV0rmhQU,1014
 gllm_inference/schema/message.pyi,sha256=VP9YppKj2mo1esl9cy6qQO9m2mMHUjTmfGDdyUor880,2220
-gllm_inference/schema/model_id.pyi,sha256=w_HA48gQ-TztufTCKS6RNlGLWGWJ9HaUyeloNdKW8sU,5816
+gllm_inference/schema/model_id.pyi,sha256=MuH0KyFjI1uC9v7PoIU6Uuk6wPdpmczVrHZj0r5EcZk,5842
 gllm_inference/schema/reasoning.pyi,sha256=SlTuiDw87GdnAn-I6YOPIJRhEBiwQljM46JohG05guQ,562
 gllm_inference/schema/token_usage.pyi,sha256=1GTQVORV0dBNmD_jix8aVaUqxMKFF04KpLP7y2urqbk,2950
 gllm_inference/schema/tool_call.pyi,sha256=zQaVxCnkVxOfOEhBidqohU85gb4PRwnwBiygKaunamk,389
@@ -131,7 +134,7 @@ gllm_inference/utils/io_utils.pyi,sha256=7kUTacHAVRYoemFUOjCH7-Qmw-YsQGd6rGYxjf_
 gllm_inference/utils/langchain.pyi,sha256=VluQiHkGigDdqLUbhB6vnXiISCP5hHqV0qokYY6dC1A,1164
 gllm_inference/utils/validation.pyi,sha256=W9RQddN90F8SJMu_HXEQyQTDMBaRL-bo7fOosZWK7oY,438
 gllm_inference.build/.gitignore,sha256=aEiIwOuxfzdCmLZe4oB1JsBmCUxwG8x-u-HBCV9JT8E,1
-gllm_inference_binary-0.5.57.dist-info/METADATA,sha256=JTadwr_lJ2tUsITlqFi2WAbqvB9M3LApoy_Xj8fbgZE,5807
-gllm_inference_binary-0.5.57.dist-info/WHEEL,sha256=GrvfTP3j0ebqecWD3AHlLRzmSrTVGeL6T8Btq6Eg9eI,108
-gllm_inference_binary-0.5.57.dist-info/top_level.txt,sha256=FpOjtN80F-qVNgbScXSEyqa0w09FYn6301iq6qt69IQ,15
-gllm_inference_binary-0.5.57.dist-info/RECORD,,
+gllm_inference_binary-0.5.65.dist-info/METADATA,sha256=ZoIOAqMSWLFjLzw3e3FTD-H3GxGsndl9lxlK4MbwsgA,5815
+gllm_inference_binary-0.5.65.dist-info/WHEEL,sha256=GrvfTP3j0ebqecWD3AHlLRzmSrTVGeL6T8Btq6Eg9eI,108
+gllm_inference_binary-0.5.65.dist-info/top_level.txt,sha256=FpOjtN80F-qVNgbScXSEyqa0w09FYn6301iq6qt69IQ,15
+gllm_inference_binary-0.5.65.dist-info/RECORD,,