gllm-inference-binary 0.5.14__cp312-cp312-manylinux_2_31_x86_64.whl → 0.5.16__cp312-cp312-manylinux_2_31_x86_64.whl

gllm_inference/builder/build_em_invoker.pyi

@@ -1,5 +1,5 @@
 from _typeshed import Incomplete
-from gllm_inference.em_invoker import AzureOpenAIEMInvoker as AzureOpenAIEMInvoker, GoogleEMInvoker as GoogleEMInvoker, LangChainEMInvoker as LangChainEMInvoker, OpenAICompatibleEMInvoker as OpenAICompatibleEMInvoker, OpenAIEMInvoker as OpenAIEMInvoker, TwelveLabsEMInvoker as TwelveLabsEMInvoker, VoyageEMInvoker as VoyageEMInvoker
+from gllm_inference.em_invoker import AzureOpenAIEMInvoker as AzureOpenAIEMInvoker, BedrockEMInvoker as BedrockEMInvoker, GoogleEMInvoker as GoogleEMInvoker, LangChainEMInvoker as LangChainEMInvoker, OpenAICompatibleEMInvoker as OpenAICompatibleEMInvoker, OpenAIEMInvoker as OpenAIEMInvoker, TwelveLabsEMInvoker as TwelveLabsEMInvoker, VoyageEMInvoker as VoyageEMInvoker
 from gllm_inference.em_invoker.em_invoker import BaseEMInvoker as BaseEMInvoker
 from gllm_inference.schema.model_id import ModelId as ModelId, ModelProvider as ModelProvider
 from typing import Any
@@ -9,6 +9,7 @@ logger: Incomplete
 
 class Key:
     """Defines valid keys in the config."""
+    ACCESS_KEY_ID: str
     API_KEY: str
     AZURE_DEPLOYMENT: str
     AZURE_ENDPOINT: str
@@ -17,6 +18,7 @@ class Key:
     MODEL_KWARGS: str
     MODEL_NAME: str
     MODEL_CLASS_PATH: str
+    SECRET_ACCESS_KEY: str
 
 def build_em_invoker(model_id: str | ModelId, credentials: str | dict[str, Any] | None = None, config: dict[str, Any] | None = None) -> BaseEMInvoker:
     '''Build an embedding model invoker based on the provided configurations.
@@ -41,6 +43,19 @@ def build_em_invoker(model_id: str | ModelId, credentials: str | dict[str, Any]
         ValueError: If the provider is invalid.
 
     Usage examples:
+        # Using Bedrock
+        ```python
+        em_invoker = build_em_invoker(
+            model_id="bedrock/cohere.embed-english-v3",
+            credentials={
+                "access_key_id": "Abc123...",
+                "secret_access_key": "Xyz123...",
+            },
+        )
+        ```
+        The credentials can also be provided through the `AWS_ACCESS_KEY_ID` and `AWS_SECRET_ACCESS_KEY`
+        environment variables.
+
         # Using Google Gen AI (via API key)
         ```python
         em_invoker = build_em_invoker(

gllm_inference/em_invoker/__init__.pyi

@@ -1,4 +1,5 @@
 from gllm_inference.em_invoker.azure_openai_em_invoker import AzureOpenAIEMInvoker as AzureOpenAIEMInvoker
+from gllm_inference.em_invoker.bedrock_em_invoker import BedrockEMInvoker as BedrockEMInvoker
 from gllm_inference.em_invoker.google_em_invoker import GoogleEMInvoker as GoogleEMInvoker
 from gllm_inference.em_invoker.langchain_em_invoker import LangChainEMInvoker as LangChainEMInvoker
 from gllm_inference.em_invoker.openai_compatible_em_invoker import OpenAICompatibleEMInvoker as OpenAICompatibleEMInvoker
@@ -6,4 +7,4 @@ from gllm_inference.em_invoker.openai_em_invoker import OpenAIEMInvoker as OpenA
 from gllm_inference.em_invoker.twelevelabs_em_invoker import TwelveLabsEMInvoker as TwelveLabsEMInvoker
 from gllm_inference.em_invoker.voyage_em_invoker import VoyageEMInvoker as VoyageEMInvoker
 
-__all__ = ['AzureOpenAIEMInvoker', 'GoogleEMInvoker', 'LangChainEMInvoker', 'OpenAIEMInvoker', 'OpenAICompatibleEMInvoker', 'TwelveLabsEMInvoker', 'VoyageEMInvoker']
+__all__ = ['AzureOpenAIEMInvoker', 'BedrockEMInvoker', 'GoogleEMInvoker', 'LangChainEMInvoker', 'OpenAIEMInvoker', 'OpenAICompatibleEMInvoker', 'TwelveLabsEMInvoker', 'VoyageEMInvoker']

gllm_inference/em_invoker/azure_openai_em_invoker.pyi

@@ -3,7 +3,7 @@ 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.em_invoker.openai_em_invoker import OpenAIEMInvoker as OpenAIEMInvoker
 from gllm_inference.em_invoker.schema.openai import Key as Key
-from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider
+from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, TruncationConfig as TruncationConfig
 from typing import Any
 
 class AzureOpenAIEMInvoker(OpenAIEMInvoker):
@@ -16,6 +16,7 @@ class AzureOpenAIEMInvoker(OpenAIEMInvoker):
         client (AsyncAzureOpenAI): The client for the Azure OpenAI API.
         default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the embedding model.
         retry_config (RetryConfig): The retry configuration for the embedding model.
+        truncation_config (TruncationConfig | None): The truncation configuration for the embedding model.
 
     Input types:
         The `AzureOpenAIEMInvoker` only supports text inputs.
@@ -68,7 +69,7 @@ class AzureOpenAIEMInvoker(OpenAIEMInvoker):
         ```
     '''
     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, retry_config: RetryConfig | 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, retry_config: RetryConfig | None = None, truncation_config: TruncationConfig | None = None) -> None:
         """Initializes a new instance of the AzureOpenAIEMInvoker class.
 
         Args:
@@ -83,4 +84,6 @@ class AzureOpenAIEMInvoker(OpenAIEMInvoker):
                 Defaults to None.
             retry_config (RetryConfig | None, optional): The retry configuration for the embedding model.
                 Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+            truncation_config (TruncationConfig | None, optional): Configuration for text truncation behavior.
+                Defaults to None, in which case no truncation is applied.
         """

gllm_inference/em_invoker/bedrock_em_invoker.pyi

@@ -0,0 +1,104 @@
+from _typeshed import Incomplete
+from enum import StrEnum
+from gllm_core.utils.retry import RetryConfig as RetryConfig
+from gllm_inference.em_invoker.em_invoker import BaseEMInvoker as BaseEMInvoker
+from gllm_inference.em_invoker.schema.bedrock import InputType as InputType, Key as Key, OutputType as OutputType
+from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, TruncationConfig as TruncationConfig, Vector as Vector
+from typing import Any
+
+class ModelType(StrEnum):
+    """Defines the type of the Bedrock embedding model."""
+    COHERE = 'cohere'
+    TITAN = 'titan'
+
+SUPPORTED_ATTACHMENTS: Incomplete
+
+class BedrockEMInvoker(BaseEMInvoker):
+    '''An embedding model invoker to interact with AWS Bedrock embedding models.
+
+    Attributes:
+        model_id (str): The model ID of the embedding model.
+        model_provider (str): The provider of the embedding model.
+        model_name (str): The name of the embedding model.
+        session (Session): The Bedrock client session.
+        client_kwargs (dict[str, Any]): The Bedrock client kwargs.
+        default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the embedding model.
+        retry_config (RetryConfig): The retry configuration for the embedding model.
+        truncation_config (TruncationConfig | None): The truncation configuration for the embedding model.
+
+    Input types:
+        The `BedrockEMInvoker` only supports text inputs.
+
+    Output format:
+        The `BedrockEMInvoker` can embed either:
+        1. A single content.
+           1. A single content is a single text.
+           2. The output will be a `Vector`, representing the embedding of the content.
+
+           # Example 1: Embedding a text content.
+           ```python
+           text = "This is a text"
+           result = await em_invoker.invoke(text)
+           ```
+
+           The above examples will return a `Vector` with a size of (embedding_size,).
+
+        2. A list of contents.
+           1. A list of contents is a list of texts.
+           2. The output will be a `list[Vector]`, where each element is a `Vector` representing the
+              embedding of each single content.
+
+           # Example: Embedding a list of contents.
+           ```python
+           text1 = "This is a text"
+           text2 = "This is another text"
+           text3 = "This is yet another text"
+           result = await em_invoker.invoke([text1, text2, text3])
+           ```
+
+           The above examples will return a `list[Vector]` with a size of (3, embedding_size).
+
+    Retry and timeout:
+        The `BedrockEMInvoker` supports retry and timeout configuration.
+        By default, the max retries is set to 0 and the timeout is set to 30.0 seconds.
+        They can be customized by providing a custom `RetryConfig` object to the `retry_config` parameter.
+
+        Retry config examples:
+        ```python
+        retry_config = RetryConfig(max_retries=0, timeout=0.0)  # No retry, no timeout
+        retry_config = RetryConfig(max_retries=0, timeout=10.0)  # No retry, 10.0 seconds timeout
+        retry_config = RetryConfig(max_retries=5, timeout=0.0)  # 5 max retries, no timeout
+        retry_config = RetryConfig(max_retries=5, timeout=10.0)  # 5 max retries, 10.0 seconds timeout
+        ```
+
+        Usage example:
+        ```python
+        em_invoker = BedrockEMInvoker(..., retry_config=retry_config)
+        ```
+    '''
+    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, retry_config: RetryConfig | None = None, truncation_config: TruncationConfig | None = None) -> None:
+        '''Initializes a new instance of the BedrockEMInvoker class.
+
+        Args:
+            model_name (str): The name of the Bedrock embedding model to be used.
+            access_key_id (str | None, optional): The AWS access key ID. Defaults to None, in which case
+                the `AWS_ACCESS_KEY_ID` environment variable will be used.
+            secret_access_key (str | None, optional): The AWS secret access key. Defaults to None, in which case
+                the `AWS_SECRET_ACCESS_KEY` environment variable will be used.
+            region_name (str, optional): The AWS region name. Defaults to "us-east-1".
+            model_kwargs (dict[str, Any] | None, optional): Additional keyword arguments for the Bedrock client.
+                Defaults to None.
+            default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the model.
+                Defaults to None.
+            retry_config (RetryConfig | None, optional): The retry configuration for the embedding model.
+                Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+            truncation_config (TruncationConfig | None, optional): Configuration for text truncation behavior.
+                Defaults to None, in which case no truncation is applied.
+
+        Raises:
+            ValueError: If the model name is not supported.
+            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.
+        '''

gllm_inference/em_invoker/em_invoker.pyi

@@ -4,7 +4,7 @@ from abc import ABC
 from gllm_core.utils.retry 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, EMContent as EMContent, ModelId as ModelId, Vector as Vector
+from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, EMContent as EMContent, ModelId as ModelId, TruncateSide as TruncateSide, TruncationConfig as TruncationConfig, Vector as Vector
 from typing import Any
 
 class BaseEMInvoker(ABC, metaclass=abc.ABCMeta):
@@ -16,12 +16,17 @@ class BaseEMInvoker(ABC, metaclass=abc.ABCMeta):
         model_id (str): The model ID of the embedding model.
         model_provider (str): The provider of the embedding model.
         model_name (str): The name of the embedding model.
-        default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the embedding model.
+        default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the
+            embedding model. Defaults to None, in which case an empty dictionary is used.
         retry_config (RetryConfig): The retry configuration for the embedding model.
+            Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+        truncation_config (TruncationConfig | None): The truncation configuration for the embedding model.
+            Defaults to None, in which case no truncation is applied.
     """
     default_hyperparameters: Incomplete
     retry_config: Incomplete
-    def __init__(self, model_id: ModelId, default_hyperparameters: dict[str, Any] | None = None, supported_attachments: set[str] | None = None, retry_config: RetryConfig | None = None) -> None:
+    truncation_config: Incomplete
+    def __init__(self, model_id: ModelId, default_hyperparameters: dict[str, Any] | None = None, supported_attachments: set[str] | None = None, retry_config: RetryConfig | None = None, truncation_config: TruncationConfig | None = None) -> None:
         """Initializes a new instance of the BaseEMInvoker class.
 
         Args:
@@ -32,6 +37,8 @@ class BaseEMInvoker(ABC, metaclass=abc.ABCMeta):
                 in which case an empty set is used (indicating that no attachments are supported).
             retry_config (RetryConfig | None, optional): The retry configuration for the embedding model.
                 Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+            truncation_config (TruncationConfig | None, optional): Configuration for text truncation behavior.
+                Defaults to None, in which case no truncation is applied.
         """
     @property
     def model_id(self) -> str:

gllm_inference/em_invoker/google_em_invoker.pyi

@@ -3,7 +3,7 @@ from gllm_core.utils.retry import RetryConfig as RetryConfig
 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
+from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, TruncationConfig as TruncationConfig, Vector as Vector
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -17,7 +17,8 @@ class GoogleEMInvoker(BaseEMInvoker):
         model_name (str): The name of the embedding model.
         client_params (dict[str, Any]): The Google client instance init parameters.
         default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the embedding model.
-        retry_config (RetryConfig | None): The retry configuration for the language model.
+        retry_config (RetryConfig): The retry configuration for the embedding model.
+        truncation_config (TruncationConfig | None): The truncation configuration for the embedding model.
 
     Initialization:
         The `GoogleEMInvoker` can use either Google Gen AI or Google Vertex AI.
@@ -98,7 +99,7 @@ class GoogleEMInvoker(BaseEMInvoker):
         ```
     '''
     client_params: 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, retry_config: RetryConfig | None = None) -> 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, retry_config: RetryConfig | None = None, truncation_config: TruncationConfig | None = None) -> None:
         '''Initializes a new instance of the GoogleEMInvoker class.
 
         Args:
@@ -117,6 +118,8 @@ class GoogleEMInvoker(BaseEMInvoker):
                 Defaults to None.
             retry_config (RetryConfig | None, optional): The retry configuration for the embedding model.
                 Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+            truncation_config (TruncationConfig | None, optional): Configuration for text truncation behavior.
+                Defaults to None, in which case no truncation is applied.
 
         Note:
             If neither `api_key` nor `credentials_path` is provided, Google Gen AI will be used by default.

gllm_inference/em_invoker/langchain_em_invoker.pyi

@@ -3,7 +3,7 @@ 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.em_invoker.em_invoker import BaseEMInvoker as BaseEMInvoker
 from gllm_inference.em_invoker.schema.langchain import Key as Key
-from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, Vector as Vector
+from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, TruncationConfig as TruncationConfig, Vector as Vector
 from gllm_inference.utils import load_langchain_model as load_langchain_model, parse_model_data as parse_model_data
 from langchain_core.embeddings import Embeddings as Embeddings
 from typing import Any
@@ -19,9 +19,10 @@ class LangChainEMInvoker(BaseEMInvoker):
         model_name (str): The name of the embedding model.
         em (Embeddings): The instance to interact with an embedding model defined using LangChain's Embeddings.
         retry_config (RetryConfig): The retry configuration for the embedding model.
+        truncation_config (TruncationConfig | None): The truncation configuration for the embedding model.
     """
     model: Incomplete
-    def __init__(self, model: Embeddings | 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, retry_config: RetryConfig | None = None) -> None:
+    def __init__(self, model: Embeddings | 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, retry_config: RetryConfig | None = None, truncation_config: TruncationConfig | None = None) -> None:
         '''Initializes a new instance of the LangChainEMInvoker class.
 
         Args:
@@ -38,4 +39,6 @@ class LangChainEMInvoker(BaseEMInvoker):
                 Defaults to None.
             retry_config (RetryConfig | None, optional): The retry configuration for the embedding model.
                 Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+            truncation_config (TruncationConfig | None, optional): Configuration for text truncation behavior.
+                Defaults to None, in which case no truncation is applied.
         '''

gllm_inference/em_invoker/openai_compatible_em_invoker.pyi

@@ -3,7 +3,7 @@ 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.em_invoker.openai_em_invoker import OpenAIEMInvoker as OpenAIEMInvoker
 from gllm_inference.em_invoker.schema.openai_compatible import Key as Key
-from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider
+from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, TruncationConfig as TruncationConfig
 from typing import Any
 
 class OpenAICompatibleEMInvoker(OpenAIEMInvoker):
@@ -16,6 +16,8 @@ class OpenAICompatibleEMInvoker(OpenAIEMInvoker):
         client (AsyncOpenAI): The OpenAI client instance.
         default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the embedding model.
         retry_config (RetryConfig): The retry configuration for the embedding model.
+        truncation_config (TruncationConfig | None): The truncation configuration for the embedding model.
+
 
     When to use:
         The `OpenAICompatibleEMInvoker` is designed to interact with endpoints that are compatible with OpenAI\'s
@@ -76,7 +78,7 @@ class OpenAICompatibleEMInvoker(OpenAIEMInvoker):
         ```
     '''
     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, retry_config: RetryConfig | 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, retry_config: RetryConfig | None = None, truncation_config: TruncationConfig | None = None) -> None:
         """Initializes a new instance of the OpenAICompatibleEMInvoker class.
 
         Args:
@@ -89,4 +91,6 @@ class OpenAICompatibleEMInvoker(OpenAIEMInvoker):
                 Defaults to None.
             retry_config (RetryConfig | None, optional): The retry configuration for the embedding model.
                 Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+            truncation_config (TruncationConfig | None, optional): Configuration for text truncation behavior.
+                Defaults to None, in which case no truncation is applied.
         """

gllm_inference/em_invoker/openai_em_invoker.pyi

@@ -4,6 +4,7 @@ from gllm_inference.constants import INVOKER_PROPAGATED_MAX_RETRIES as INVOKER_P
 from gllm_inference.em_invoker.em_invoker import BaseEMInvoker as BaseEMInvoker
 from gllm_inference.em_invoker.schema.openai import Key as Key
 from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, Vector as Vector
+from gllm_inference.schema.config import TruncationConfig as TruncationConfig
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -18,6 +19,7 @@ class OpenAIEMInvoker(BaseEMInvoker):
         client (AsyncOpenAI): The client for the OpenAI API.
         default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the embedding model.
         retry_config (RetryConfig): The retry configuration for the embedding model.
+        truncation_config (TruncationConfig | None): The truncation configuration for the embedding model.
 
     Input types:
         The `OpenAIEMInvoker` only supports text inputs.
@@ -70,7 +72,7 @@ class OpenAIEMInvoker(BaseEMInvoker):
         ```
     '''
     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, retry_config: RetryConfig | None = None) -> 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, retry_config: RetryConfig | None = None, truncation_config: TruncationConfig | None = None) -> None:
         """Initializes a new instance of the OpenAIEMInvoker class.
 
         Args:
@@ -83,4 +85,6 @@ class OpenAIEMInvoker(BaseEMInvoker):
                 Defaults to None.
             retry_config (RetryConfig | None, optional): The retry configuration for the embedding model.
                 Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+            truncation_config (TruncationConfig | None, optional): Configuration for text truncation behavior.
+                Defaults to None, in which case no truncation is applied.
         """

gllm_inference/em_invoker/schema/bedrock.pyi

@@ -0,0 +1,20 @@
+class Key:
+    """Defines valid keys in Bedrock."""
+    ACCEPT: str
+    CONTENT_TYPE: str
+    INPUT_TEXT: str
+    INPUT_TYPE: str
+    MODEL_ID: str
+    TEXTS: str
+
+class InputType:
+    """Defines valid input types in Bedrock."""
+    APPLICATION_JSON: str
+    SEARCH_DOCUMENT: str
+    SEARCH_QUERY: str
+
+class OutputType:
+    """Defines valid output types in Bedrock."""
+    BODY: str
+    EMBEDDING: str
+    EMBEDDINGS: str

gllm_inference/em_invoker/twelevelabs_em_invoker.pyi

@@ -3,7 +3,7 @@ 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.em_invoker.em_invoker import BaseEMInvoker as BaseEMInvoker
 from gllm_inference.em_invoker.schema.twelvelabs import InputType as InputType, Key as Key, OutputType as OutputType
-from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, EMContent as EMContent, ModelId as ModelId, ModelProvider as ModelProvider, Vector as Vector
+from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, EMContent as EMContent, ModelId as ModelId, ModelProvider as ModelProvider, TruncationConfig as TruncationConfig, Vector as Vector
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -18,6 +18,7 @@ class TwelveLabsEMInvoker(BaseEMInvoker):
         client (Client): The client for the TwelveLabs API.
         default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the embedding model.
         retry_config (RetryConfig): The retry configuration for the embedding model.
+        truncation_config (TruncationConfig | None): The truncation configuration for the embedding model.
 
     Input types:
         The `TwelveLabsEMInvoker` supports the following input types: text, audio, and image.
@@ -83,7 +84,7 @@ class TwelveLabsEMInvoker(BaseEMInvoker):
         ```
     '''
     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, retry_config: RetryConfig | None = None) -> 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, retry_config: RetryConfig | None = None, truncation_config: TruncationConfig | None = None) -> None:
         """Initializes a new instance of the TwelveLabsEMInvoker class.
 
         Args:
@@ -96,4 +97,6 @@ class TwelveLabsEMInvoker(BaseEMInvoker):
                 Defaults to None.
             retry_config (RetryConfig | None, optional): The retry configuration for the embedding model.
                 Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+            truncation_config (TruncationConfig | None, optional): Configuration for text truncation behavior.
+                Defaults to None, in which case no truncation is applied.
         """

gllm_inference/em_invoker/voyage_em_invoker.pyi

@@ -3,7 +3,7 @@ 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.em_invoker.em_invoker import BaseEMInvoker as BaseEMInvoker
 from gllm_inference.em_invoker.schema.voyage import InputType as InputType, Key as Key
-from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, EMContent as EMContent, ModelId as ModelId, ModelProvider as ModelProvider, Vector as Vector
+from gllm_inference.schema import Attachment as Attachment, AttachmentType as AttachmentType, EMContent as EMContent, ModelId as ModelId, ModelProvider as ModelProvider, TruncationConfig as TruncationConfig, Vector as Vector
 from typing import Any
 
 SUPPORTED_ATTACHMENTS: Incomplete
@@ -19,6 +19,7 @@ class VoyageEMInvoker(BaseEMInvoker):
         client (Client): The client for the Voyage API.
         default_hyperparameters (dict[str, Any]): Default hyperparameters for invoking the embedding model.
         retry_config (RetryConfig): The retry configuration for the embedding model.
+        truncation_config (TruncationConfig | None): The truncation configuration for the embedding model.
 
     Input types:
         The `VoyageEMInvoker` supports the following input types: text, image, and a tuple containing text and image.
@@ -85,7 +86,7 @@ class VoyageEMInvoker(BaseEMInvoker):
         ```
     '''
     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, retry_config: RetryConfig | None = None) -> 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, retry_config: RetryConfig | None = None, truncation_config: TruncationConfig | None = None) -> None:
         """Initializes a new instance of the VoyageEMInvoker class.
 
         Args:
@@ -98,4 +99,6 @@ class VoyageEMInvoker(BaseEMInvoker):
                 Defaults to None.
             retry_config (RetryConfig | None, optional): The retry configuration for the embedding model.
                 Defaults to None, in which case a default config with no retry and 30.0 seconds timeout will be used.
+            truncation_config (TruncationConfig | None, optional): Configuration for text truncation behavior.
+                Defaults to None, in which case no truncation is applied.
         """

gllm_inference/schema/__init__.pyi

@@ -1,6 +1,7 @@
 from gllm_inference.schema.attachment import Attachment as Attachment
 from gllm_inference.schema.code_exec_result import CodeExecResult as CodeExecResult
-from gllm_inference.schema.enums import AttachmentType as AttachmentType, EmitDataType as EmitDataType, MessageRole as MessageRole
+from gllm_inference.schema.config import TruncationConfig as TruncationConfig
+from gllm_inference.schema.enums import AttachmentType as AttachmentType, EmitDataType as EmitDataType, MessageRole as MessageRole, TruncateSide as TruncateSide
 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
@@ -10,4 +11,4 @@ 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, ErrorResponse as ErrorResponse, MessageContent as MessageContent, ResponseSchema as ResponseSchema, Vector as Vector
 
-__all__ = ['Attachment', 'AttachmentType', 'CodeExecResult', 'EMContent', 'EmitDataType', 'ErrorResponse', 'InputTokenDetails', 'MessageContent', 'LMOutput', 'ModelId', 'ModelProvider', 'Message', 'MessageRole', 'OutputTokenDetails', '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', 'TruncateSide', 'TruncationConfig', 'Vector']

gllm_inference/schema/config.pyi

@@ -0,0 +1,15 @@
+from gllm_inference.schema.enums import TruncateSide as TruncateSide
+from pydantic import BaseModel
+
+class TruncationConfig(BaseModel):
+    """Configuration for text truncation behavior.
+
+    Attributes:
+        max_length (int): Maximum length of text content. Required.
+        truncate_side (TruncateSide | None): Side to truncate from when max_length is exceeded.
+            1. TruncateSide.RIGHT: Keep the beginning of the text, truncate from the end (default)
+            2. TruncateSide.LEFT: Keep the end of the text, truncate from the beginning
+            If None, defaults to TruncateSide.RIGHT
+    """
+    max_length: int
+    truncate_side: TruncateSide | None

gllm_inference/schema/enums.pyi

@@ -22,3 +22,8 @@ class MessageRole(StrEnum):
     SYSTEM = 'system'
     USER = 'user'
     ASSISTANT = 'assistant'
+
+class TruncateSide(StrEnum):
+    """Enumeration for truncation sides."""
+    RIGHT = 'RIGHT'
+    LEFT = 'LEFT'

gllm_inference.pyi

@@ -13,6 +13,7 @@ import typing
 import gllm_core
 import gllm_core.utils
 import gllm_inference.em_invoker.AzureOpenAIEMInvoker
+import gllm_inference.em_invoker.BedrockEMInvoker
 import gllm_inference.em_invoker.GoogleEMInvoker
 import gllm_inference.em_invoker.LangChainEMInvoker
 import gllm_inference.em_invoker.OpenAICompatibleEMInvoker
@@ -41,17 +42,21 @@ import gllm_inference.request_processor.LMRequestProcessor
 import gllm_core.utils.imports
 import gllm_inference.schema.ModelId
 import gllm_inference.schema.ModelProvider
+import gllm_inference.schema.TruncationConfig
 import openai
+import asyncio
+import enum
+import gllm_inference.schema.Vector
+import aioboto3
 import gllm_inference.exceptions.parse_error_message
 import gllm_inference.schema.Attachment
 import gllm_inference.schema.AttachmentType
 import gllm_inference.schema.EMContent
-import gllm_inference.schema.Vector
+import gllm_inference.schema.TruncateSide
 import google
 import google.auth
 import google.genai
 import google.genai.types
-import asyncio
 import concurrent
 import concurrent.futures
 import concurrent.futures.ThreadPoolExecutor
@@ -68,7 +73,6 @@ import voyageai
 import voyageai.client_async
 import asyncio.CancelledError
 import asyncio.TimeoutError
-import enum
 import http
 import http.HTTPStatus
 import aiohttp
@@ -88,7 +92,6 @@ import gllm_inference.schema.TokenUsage
 import gllm_inference.schema.ToolCall
 import gllm_inference.schema.ToolResult
 import anthropic
-import aioboto3
 import gllm_inference.schema.MessageRole
 import langchain_core.language_models
 import langchain_core.messages

{gllm_inference_binary-0.5.14.dist-info → gllm_inference_binary-0.5.16.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gllm-inference-binary
-Version: 0.5.14
+Version: 0.5.16
 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-0.5.14.dist-info → gllm_inference_binary-0.5.16.dist-info}/RECORD

@@ -1,6 +1,6 @@
 gllm_inference/__init__.pyi,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 gllm_inference/builder/__init__.pyi,sha256=usz2lvfwO4Yk-ZGKXbCWG1cEr3nlQXxMNDNC-2yc1NM,500
-gllm_inference/builder/build_em_invoker.pyi,sha256=YL71GriZEXn4uxmXBJHWC200QdWRPwUJY_G0kKi5-dk,5352
+gllm_inference/builder/build_em_invoker.pyi,sha256=7JTt8XpfqLQdv5cltIyi1r6Sf8MAUKgornHn5z57raA,5873
 gllm_inference/builder/build_lm_invoker.pyi,sha256=igJdLnWiY5QcdT4EClvgxFGSVZjARpq5hz2FYcBHWEQ,6876
 gllm_inference/builder/build_lm_request_processor.pyi,sha256=33Gi3onftl-V2e_mkJios5zmXRKSoAVPX3UK7YBExjk,4491
 gllm_inference/builder/build_output_parser.pyi,sha256=_Lrq-bh1oPsb_Nwkkr_zyEUwIOMysRFZkvEtEM29LZM,936
@@ -9,24 +9,26 @@ gllm_inference/catalog/catalog.pyi,sha256=a4RNG1lKv51GxQpOqh47tz-PAROMPaeP2o5XNL
 gllm_inference/catalog/lm_request_processor_catalog.pyi,sha256=ranHMbG9--DZj9FJRhIUa6U8e-L-Tm-_hSBpzJ6DDs4,5428
 gllm_inference/catalog/prompt_builder_catalog.pyi,sha256=OU8k_4HbqjZEzHZlzSM3uzGQZJmM2uGD76Csqom0CEQ,3197
 gllm_inference/constants.pyi,sha256=KDzHmjVYjd0uTHqLgIzgHhLZx99D8jGBIg74eO8KQv0,314
-gllm_inference/em_invoker/__init__.pyi,sha256=XESsrYo1PZeeHe7AMRyuzKoV7XDD5oN89ZTH01zRf4k,873
-gllm_inference/em_invoker/azure_openai_em_invoker.pyi,sha256=OEkVu5nv92ITqdhDtgDg4MiLSDRWDmLSnAhYtXpCn6E,4602
-gllm_inference/em_invoker/em_invoker.pyi,sha256=hiH8FB5R-KxhI8Ds2htF3cjRcIcH92yHPcOdpgc4FDo,4341
-gllm_inference/em_invoker/google_em_invoker.pyi,sha256=vj0KAX5b2rhvqqbpjFZYLmk84RHuNw0pXkhx3bbHmM4,6182
+gllm_inference/em_invoker/__init__.pyi,sha256=83QVCkMjS2-jMKdAvmZska4LuJ-un755lAxjuVSLZ9o,987
+gllm_inference/em_invoker/azure_openai_em_invoker.pyi,sha256=eJd1Ygths8wu0w2S9VmF-2iEQj5zs7nACyNJXAsNXUA,4983
+gllm_inference/em_invoker/bedrock_em_invoker.pyi,sha256=KdX1PMWPfrmFIrSbf8Y6jkDuwc81Q9XWR9DQPLoO66I,5456
+gllm_inference/em_invoker/em_invoker.pyi,sha256=dgIeAIetQcBmohwYwgo1vNw7YNO_3DQCobUaabBtf7g,5043
+gllm_inference/em_invoker/google_em_invoker.pyi,sha256=oDS4dXBcLg59ePeiLTwdl09927oJNZ_ykIe0n6Ba8gU,6557
 gllm_inference/em_invoker/langchain/__init__.pyi,sha256=VYGKE5OgU0my1RlhgzkU_A7-GLGnUDDnNFuctuRwILE,148
 gllm_inference/em_invoker/langchain/em_invoker_embeddings.pyi,sha256=6nASLqi0FXCpqyYPl7kM3g7hAW-xS5ZwsS3GFudns98,2347
-gllm_inference/em_invoker/langchain_em_invoker.pyi,sha256=lXiTTGcNOIwurZx3_6vWLE1DQyioK8Z1fOcPuUATnxA,2782
-gllm_inference/em_invoker/openai_compatible_em_invoker.pyi,sha256=Qz2Qx1KRKhzXr8IseDWcF_6yC-SNtfsXvQuGuKnqVe8,4978
-gllm_inference/em_invoker/openai_em_invoker.pyi,sha256=SFuS2DsvMHcibxFnpQOOchlZUyNRRlI2uMhVEUfifas,4235
+gllm_inference/em_invoker/langchain_em_invoker.pyi,sha256=HuQD5Do4jwqKoMMgMjgZkic2L21n2ayJewIce09bZ3M,3163
+gllm_inference/em_invoker/openai_compatible_em_invoker.pyi,sha256=GudWfL7QCAKLInMw94jTNogbyELlD9tDbrDErHB4RRI,5360
+gllm_inference/em_invoker/openai_em_invoker.pyi,sha256=vsfEmDNvwrlBhDxqsCKyMpMZbl_FaQUWHEgQc9yeo14,4656
 gllm_inference/em_invoker/schema/__init__.pyi,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+gllm_inference/em_invoker/schema/bedrock.pyi,sha256=sqPHG0jBc09K4V1aSpybqhqIdYfZQTB2VD_nJ9j_ttQ,423
 gllm_inference/em_invoker/schema/google.pyi,sha256=MUmgtjMmjSpzmzaAOx6JGZbcdRxgMUhOpvcVQIo-oGs,146
 gllm_inference/em_invoker/schema/langchain.pyi,sha256=onpZutqa2xw2g8rdJTdycy3ub58lkPBVB3KvVVPpyds,92
 gllm_inference/em_invoker/schema/openai.pyi,sha256=Q_dsEcodkOXYXPdrkOkW0LnuLhfeq8tEbtZAGMz2ajA,139
 gllm_inference/em_invoker/schema/openai_compatible.pyi,sha256=gmvGtsWoOMBelke_tZjC6dKimFBW9f4Vrgv0Ig0OM9Q,150
 gllm_inference/em_invoker/schema/twelvelabs.pyi,sha256=F6wKHgG01bYskJpKoheBSpRpHUfFpteKn9sj9n5YfcU,372
 gllm_inference/em_invoker/schema/voyage.pyi,sha256=HVpor0fqNy-IwapCICfsgFmqf1FJXCOMIxS2vOXhHd8,289
-gllm_inference/em_invoker/twelevelabs_em_invoker.pyi,sha256=6b8oJ8aLNBn45o85dka-Xbq9nWZqtcMgFls39ekEZ6o,5066
-gllm_inference/em_invoker/voyage_em_invoker.pyi,sha256=7akaf8GxOA8Trokad0xmlYKr49rY3Egm3_4gpW-vly8,5134
+gllm_inference/em_invoker/twelevelabs_em_invoker.pyi,sha256=RYY4td4N3cPSy-yXTEUKwYLz0AH8mLxi2j5Bfr9PS0g,5447
+gllm_inference/em_invoker/voyage_em_invoker.pyi,sha256=zJZqMvvFKu3sHdrNM773UYjfHVlnwE2w2BmvdFcHzV0,5515
 gllm_inference/exceptions/__init__.pyi,sha256=v9uxjW5DssIn7n_bKqT7L83CeqFET2Z45GFOvi78UuE,977
 gllm_inference/exceptions/error_parser.pyi,sha256=4RkVfS2Fl9kjz_h2bK9eoAeI-Y-VkHcUqXWj68BsYig,2393
 gllm_inference/exceptions/exceptions.pyi,sha256=5YRackwVNvyOJjOtiVszqu8q87s8ioXTa-XwaYmeiC4,4643
@@ -76,10 +78,11 @@ gllm_inference/prompt_formatter/prompt_formatter.pyi,sha256=UkcPi5ao98OGJyNRsqfh
 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=He-ytjwv2H5Hn312WFBAlBK96ALKTtDO3AT_80hCGTg,2321
-gllm_inference/schema/__init__.pyi,sha256=aQx7eKVsFNgJkYtbbKppa9Ww2WfTe0yzl_ewajDCo6w,1363
+gllm_inference/schema/__init__.pyi,sha256=Kt2ei_dr7_66p-Y1Rzum19mYldR0CLKJAJZiEEb9KRM,1507
 gllm_inference/schema/attachment.pyi,sha256=jApuzjOHJDCz4lr4MlHzBgIndh559nbWu2Xp1fk3hso,3297
 gllm_inference/schema/code_exec_result.pyi,sha256=ZTHh6JtRrPIdQ059P1UAiD2L-tAO1_S5YcMsAXfJ5A0,559
-gllm_inference/schema/enums.pyi,sha256=Nvc_Qsd8yyiH_tCtymN39O6EZ0DT_wxYHmlKLllxC00,605
+gllm_inference/schema/config.pyi,sha256=rAL_UeXyQeXVk1P2kqd8vFWOMwmKenfpQLtvMP74t9s,674
+gllm_inference/schema/enums.pyi,sha256=w5Bq3m-Ixl4yAd4801APhw9fjCiuqttWuUXWvSWSEEs,717
 gllm_inference/schema/lm_output.pyi,sha256=GafJV0KeD-VSwWkwG1oz-uruXrQ7KDZTuoojPCBRpg8,1956
 gllm_inference/schema/message.pyi,sha256=VP9YppKj2mo1esl9cy6qQO9m2mMHUjTmfGDdyUor880,2220
 gllm_inference/schema/model_id.pyi,sha256=0UJS7M91hPlzWuZI3CMGZm9ewYrTxmLUMtIhHqJOg0Q,5481
@@ -92,8 +95,8 @@ gllm_inference/utils/__init__.pyi,sha256=npmBmmlBv7cPHMg1hdL3S2_RelD6vk_LhCsGELh
 gllm_inference/utils/langchain.pyi,sha256=VluQiHkGigDdqLUbhB6vnXiISCP5hHqV0qokYY6dC1A,1164
 gllm_inference/utils/validation.pyi,sha256=toxBtRp-VItC_X7sNi-GDd7sjibBdWMrR0q01OI2D7k,385
 gllm_inference.build/.gitignore,sha256=aEiIwOuxfzdCmLZe4oB1JsBmCUxwG8x-u-HBCV9JT8E,1
-gllm_inference.cpython-312-x86_64-linux-gnu.so,sha256=5ZI_ouZxclillOnEuYy3w5yrlRnu3x1iCRSes08hgJQ,4215752
-gllm_inference.pyi,sha256=Ho74MJb64Q0MQI7xTOf7pmM675HXs3ZQfLKg2kmusDg,3520
-gllm_inference_binary-0.5.14.dist-info/METADATA,sha256=_PX1mmZ2KxIQu7Tp9zlYLjTwSBl0UzG9ewo8UPe5y6w,4608
-gllm_inference_binary-0.5.14.dist-info/WHEEL,sha256=mNY4pwQL4AOAoPmLYEQs2SSpMIbATFeiJFktRD5iKkY,110
-gllm_inference_binary-0.5.14.dist-info/RECORD,,
+gllm_inference.cpython-312-x86_64-linux-gnu.so,sha256=g-1RtGrQtFBRD143QzVGqSetkn9kayawjat7sm5cgro,4334728
+gllm_inference.pyi,sha256=0PMbN8u5rnM8r9fZQFDM9V_UuvlYu3fpX6iLH4NKioA,3658
+gllm_inference_binary-0.5.16.dist-info/METADATA,sha256=Hjc4sS2zCgCUwhPEkaUTql8XLPjSdAArBYoTFKqv-nE,4608
+gllm_inference_binary-0.5.16.dist-info/WHEEL,sha256=mNY4pwQL4AOAoPmLYEQs2SSpMIbATFeiJFktRD5iKkY,110
+gllm_inference_binary-0.5.16.dist-info/RECORD,,