gllm-inference-binary 0.4.62__cp311-cp311-manylinux_2_31_x86_64.whl → 0.5.1__cp311-cp311-manylinux_2_31_x86_64.whl

gllm_inference/builder/build_lm_request_processor.pyi

@@ -1,7 +1,6 @@
 from _typeshed import Incomplete
 from gllm_inference.builder.build_lm_invoker import build_lm_invoker as build_lm_invoker
 from gllm_inference.builder.build_output_parser import build_output_parser as build_output_parser
-from gllm_inference.output_parser.output_parser import BaseOutputParser as BaseOutputParser
 from gllm_inference.prompt_builder import PromptBuilder as PromptBuilder
 from gllm_inference.request_processor.lm_request_processor import LMRequestProcessor as LMRequestProcessor
 from gllm_inference.schema.model_id import ModelId as ModelId
@@ -9,7 +8,7 @@ from typing import Any
 
 logger: Incomplete
 
-def build_lm_request_processor(model_id: str | ModelId, credentials: str | dict[str, Any] | None = None, config: dict[str, Any] | None = None, system_template: str = '', user_template: str = '', output_parser_type: str = 'none', output_parser: BaseOutputParser | None = None) -> LMRequestProcessor:
+def build_lm_request_processor(model_id: str | ModelId, credentials: str | dict[str, Any] | None = None, config: dict[str, Any] | None = None, system_template: str = '', user_template: str = '', output_parser_type: str = 'none') -> LMRequestProcessor:
     '''Build a language model invoker based on the provided configurations.
 
     Args:
@@ -32,8 +31,6 @@ def build_lm_request_processor(model_id: str | ModelId, credentials: str | dict[
             Defaults to an empty string.
         output_parser_type (str, optional): The type of output parser to use. Supports "json" and "none".
             Defaults to "none".
-        output_parser (BaseOutputParser | None, optional): Deprecated parameter to pass an output parser.
-            Will be removed in v0.5.0. Defaults to None.
 
     Returns:
         LMRequestProcessor: The initialized language model request processor.

gllm_inference/catalog/catalog.pyi

@@ -13,46 +13,42 @@ class BaseCatalog(ABC, BaseModel, Generic[T], arbitrary_types_allowed=True, meta
     Attributes:
         components (dict[str, T]): A dictionary containing the components.
 
-    Load from Google Sheets using client email and private key example:
-    ```python
-    catalog = BaseCatalog.from_gsheets(
-        sheet_id="...",
-        worksheet_id="...",
-        client_email="...",
-        private_key="...",
-    )
-
-    component = catalog.name
-    ```
-
-    Load from Google Sheets using credential file example:
-    ```python
-    catalog = BaseCatalog.from_gsheets(
-        sheet_id="...",
-        worksheet_id="...",
-        credential_file_path="...",
-    )
-
-    component = catalog.name
-    ```
-
-    Load from CSV example:
-    ```python
-    catalog = BaseCatalog.from_csv(csv_path="...")
-
-    component = catalog.name
-    ```
-
-    Load from records example:
-    ```python
-    catalog = BaseCatalog.from_records(
-        records=[
+    Initialization:
+        # Example 1: Load from Google Sheets using client email and private key
+        ```python
+        catalog = BaseCatalog.from_gsheets(
+            sheet_id="...",
+            worksheet_id="...",
+            client_email="...",
+            private_key="...",
+        )
+        component = catalog.name
+        ```
+
+        # Example 2: Load from Google Sheets using credential file
+        ```python
+        catalog = BaseCatalog.from_gsheets(
+            sheet_id="...",
+            worksheet_id="...",
+            credential_file_path="...",
+        )
+        component = catalog.name
+        ```
+
+        # Example 3: Load from CSV
+        ```python
+        catalog = BaseCatalog.from_csv(csv_path="...")
+        component = catalog.name
+        ```
+
+        # Example 4: Load from records
+        ```python
+        records = [
             {"name": "...", "col_1": "...", "col_2": "..."},
             {"name": "...", "col_1": "...", "col_2": "..."},
-        ],
-    )
-
-    component = catalog.name
+        ]
+        catalog = BaseCatalog.from_records(records=records)
+        component = catalog.name
     ```
     '''
     components: dict[str, T]

gllm_inference/catalog/lm_request_processor_catalog.pyi

@@ -1,12 +1,6 @@
 from _typeshed import Incomplete
-from gllm_inference.builder import build_lm_request_processor as build_lm_request_processor, build_output_parser as build_output_parser
-from gllm_inference.catalog.catalog import BaseCatalog as BaseCatalog, logger as logger
-from gllm_inference.catalog.component_map import LM_INVOKER_TYPE_MAP as LM_INVOKER_TYPE_MAP, PROMPT_BUILDER_TYPE_MAP as PROMPT_BUILDER_TYPE_MAP
-from gllm_inference.lm_invoker.lm_invoker import BaseLMInvoker as BaseLMInvoker
-from gllm_inference.multimodal_lm_invoker.multimodal_lm_invoker import BaseMultimodalLMInvoker as BaseMultimodalLMInvoker
-from gllm_inference.multimodal_prompt_builder.multimodal_prompt_builder import MultimodalPromptBuilder as MultimodalPromptBuilder
-from gllm_inference.output_parser.output_parser import BaseOutputParser as BaseOutputParser
-from gllm_inference.prompt_builder.prompt_builder import BasePromptBuilder as BasePromptBuilder, PromptBuilder as PromptBuilder
+from gllm_inference.builder import build_lm_request_processor as build_lm_request_processor
+from gllm_inference.catalog.catalog import BaseCatalog as BaseCatalog
 from gllm_inference.request_processor import LMRequestProcessor as LMRequestProcessor
 
 MODEL_ID_ENV_VAR_REGEX_PATTERN: str
@@ -67,8 +61,8 @@ class LMRequestProcessorCatalog(BaseCatalog[LMRequestProcessor]):
 
     Template Format Example:
         # Example 1: Google Sheets
-        For an example of how a Google Sheets file can be formatted to be loaded using LMRequestProcessorCatalog,
-        see: https://docs.google.com/spreadsheets/d/1CX9i45yEinv1UdB3s6uHNMj7mxr2-s1NFHfFDvMsq0E/edit?usp=drive_link
+        For an example of how a Google Sheets file can be formatted to be loaded using LMRequestProcessorCatalog, see:
+        https://docs.google.com/spreadsheets/d/1CX9i45yEinv1UdB3s6uHNMj7mxr2-s1NFHfFDvMsq0E/edit?usp=drive_link
 
         # Example 2: CSV
         For an example of how a CSV file can be formatted to be loaded using LMRequestProcessorCatalog, see:

gllm_inference/catalog/prompt_builder_catalog.pyi

@@ -1,18 +1,14 @@
 from _typeshed import Incomplete
-from gllm_inference.catalog.catalog import BaseCatalog as BaseCatalog, logger as logger
-from gllm_inference.catalog.component_map import PROMPT_BUILDER_TYPE_MAP as PROMPT_BUILDER_TYPE_MAP
-from gllm_inference.multimodal_prompt_builder import MultimodalPromptBuilder as MultimodalPromptBuilder
-from gllm_inference.prompt_builder.prompt_builder import BasePromptBuilder as BasePromptBuilder, PromptBuilder as PromptBuilder
+from gllm_inference.catalog.catalog import BaseCatalog as BaseCatalog
+from gllm_inference.prompt_builder.prompt_builder import PromptBuilder as PromptBuilder
 
-PROMPT_BUILDER_MODEL_PARAM_MAP: Incomplete
 PROMPT_BUILDER_REQUIRED_COLUMNS: Incomplete
 
-class PromptBuilderCatalog(BaseCatalog[BasePromptBuilder | MultimodalPromptBuilder | PromptBuilder]):
+class PromptBuilderCatalog(BaseCatalog[PromptBuilder]):
     '''Loads multiple prompt builders from certain sources.
 
     Attributes:
-        components (dict[str, BasePromptBuilder | MultimodalPromptBuilder | PromptBuilder]):
-            Dictionary of the loaded prompt builders.
+        components (dict[str, PromptBuilder]): Dictionary of the loaded prompt builders.
 
     Initialization:
         # Example 1: Load from Google Sheets using client email and private key
@@ -23,7 +19,6 @@ class PromptBuilderCatalog(BaseCatalog[BasePromptBuilder | MultimodalPromptBuild
             client_email="...",
             private_key="...",
         )
-
         prompt_builder = catalog.name
         ```
 
@@ -34,42 +29,38 @@ class PromptBuilderCatalog(BaseCatalog[BasePromptBuilder | MultimodalPromptBuild
             worksheet_id="...",
             credential_file_path="...",
         )
-
         prompt_builder = catalog.name
         ```
 
         # Example 3: Load from CSV
         ```python
         catalog = PromptBuilderCatalog.from_csv(csv_path="...")
-
         prompt_builder = catalog.name
         ```
 
         # Example 4: Load from records
         ```python
-        catalog = PromptBuilderCatalog.from_records(
-            records=[
-                {
-                    "name": "summarize",
-                    "system": "You are an AI expert\\nSummarize the following context.\\n\\nContext:\\n```{context}```",
-                    "user": ""
-                },
-                {
-                    "name": "transform_query",
-                    "system": "",
-                    "user": "Transform the following query into a simpler form.\\n\\nQuery:\\n```{query}```"
-                },
-                {
-                    "name": "draft_document",
-                    "system": (
-                        "You are an AI expert.\\nDraft a document following the provided format and context.\\n\\n"
-                        "Format:\\n```{format}```\\n\\nContext:\\n```{context}```"
-                    ),
-                    "user": "User instruction:\\n{query}"
-                },
-            ]
-        )
-
+        records=[
+            {
+                "name": "summarize",
+                "system": "You are an AI expert\\nSummarize the following context.\\n\\nContext:\\n```{context}```",
+                "user": ""
+            },
+            {
+                "name": "transform_query",
+                "system": "",
+                "user": "Transform the following query into a simpler form.\\n\\nQuery:\\n```{query}```"
+            },
+            {
+                "name": "draft_document",
+                "system": (
+                    "You are an AI expert.\\nDraft a document following the provided format and context.\\n\\n"
+                    "Format:\\n```{format}```\\n\\nContext:\\n```{context}```"
+                ),
+                "user": "User instruction:\\n{query}"
+            },
+        ]
+        catalog = PromptBuilderCatalog.from_records(records=records)
         prompt_builder = catalog.answer_question
         ```
 
@@ -91,7 +82,4 @@ class PromptBuilderCatalog(BaseCatalog[BasePromptBuilder | MultimodalPromptBuild
 
         Important Notes:
             1. At least one of the `system` and `user` columns must be filled.
-
-    WARNING: The use of BasePromptBuilder | MultimodalPromptBuilder is deprecated and will be removed in version 0.5.0.
-    Please use PromptBuilder instead.
     '''

gllm_inference/constants.pyi

@@ -1,10 +1,7 @@
 from _typeshed import Incomplete
 
-ALL_EXTENSIONS: str
 DEFAULT_AZURE_OPENAI_API_VERSION: str
 DOCUMENT_MIME_TYPES: Incomplete
 GOOGLE_SCOPES: Incomplete
 HEX_REPR_LENGTH: int
-MESSAGE_TUPLE_LENGTH: int
-DEFAULT_CONTENT_PLACEHOLDER_TYPE: str
 HTTP_STATUS_CODE_PATTERNS: Incomplete

gllm_inference/em_invoker/__init__.pyi

@@ -1,12 +1,9 @@
 from gllm_inference.em_invoker.azure_openai_em_invoker import AzureOpenAIEMInvoker as AzureOpenAIEMInvoker
 from gllm_inference.em_invoker.google_em_invoker import GoogleEMInvoker as GoogleEMInvoker
-from gllm_inference.em_invoker.google_generativeai_em_invoker import GoogleGenerativeAIEMInvoker as GoogleGenerativeAIEMInvoker
-from gllm_inference.em_invoker.google_vertexai_em_invoker import GoogleVertexAIEMInvoker as GoogleVertexAIEMInvoker
 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
 from gllm_inference.em_invoker.openai_em_invoker import OpenAIEMInvoker as OpenAIEMInvoker
-from gllm_inference.em_invoker.tei_em_invoker import TEIEMInvoker as TEIEMInvoker
 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', 'GoogleGenerativeAIEMInvoker', 'GoogleVertexAIEMInvoker', 'LangChainEMInvoker', 'OpenAIEMInvoker', 'OpenAICompatibleEMInvoker', 'TEIEMInvoker', 'TwelveLabsEMInvoker', 'VoyageEMInvoker']
+__all__ = ['AzureOpenAIEMInvoker', 'GoogleEMInvoker', 'LangChainEMInvoker', 'OpenAIEMInvoker', 'OpenAICompatibleEMInvoker', 'TwelveLabsEMInvoker', 'VoyageEMInvoker']

gllm_inference/em_invoker/em_invoker.pyi

@@ -2,14 +2,11 @@ import abc
 from _typeshed import Incomplete
 from abc import ABC
 from gllm_core.utils.retry import RetryConfig
-from gllm_inference.constants import ALL_EXTENSIONS as ALL_EXTENSIONS, DOCUMENT_MIME_TYPES as DOCUMENT_MIME_TYPES
+from gllm_inference.constants import DOCUMENT_MIME_TYPES as DOCUMENT_MIME_TYPES
 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 langchain_core.embeddings import Embeddings as Embeddings
 from typing import Any
 
-DEPRECATED_DETAILS: str
-
 class BaseEMInvoker(ABC, metaclass=abc.ABCMeta):
     """A base class for embedding model invokers used in Gen AI applications.
 
@@ -24,22 +21,18 @@ class BaseEMInvoker(ABC, metaclass=abc.ABCMeta):
     """
     default_hyperparameters: Incomplete
     retry_config: Incomplete
-    def __init__(self, model_id: ModelId, default_hyperparameters: dict[str, Any] | None = None, valid_extensions_map: dict[str, set[str]] | None = None, retry_config: RetryConfig | None = None, langchain_kwargs: dict[str, Any] | None = None) -> None:
-        '''Initializes a new instance of the BaseEMInvoker class.
+    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:
+        """Initializes a new instance of the BaseEMInvoker class.
 
         Args:
             model_id (ModelId): The model ID of the embedding model.
             default_hyperparameters (dict[str, Any] | None, optional): Default hyperparameters for invoking the
                 embedding model. Defaults to None, in which case an empty dictionary is used.
-            valid_extensions_map (dict[str, set[str]] | None, optional): A dictionary mapping for validating the
-                content type of the multimodal inputs. They keys are the mime types (e.g. "image") and the values are
-                the set of valid file extensions for the corresponding mime type. 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).
             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 is used.
-            langchain_kwargs (dict[str, Any] | None, optional): Additional keyword arguments to initiate the LangChain
-                embedding model. Defaults to None.
-        '''
+        """
     @property
     def model_id(self) -> str:
         """The model ID of the embedding model.
@@ -88,15 +81,3 @@ class BaseEMInvoker(ABC, metaclass=abc.ABCMeta):
             TimeoutError: If the invocation times out.
             ValueError: If the input content is invalid.
         """
-    def to_langchain(self) -> Embeddings:
-        """Converts the current embedding model invoker to an instance of LangChain `Embeddings` object.
-
-        This method converts the EM invoker to an instance of LangChain's `Embeddings` object.
-        This method requires the appropriate `langchain-<provider>` package to be installed.
-
-        Returns:
-            Embeddings: An instance of LangChain `Embeddings` object.
-
-        Raises:
-            ValueError: If `langchain_module_name` or `langchain_class_name` is missing.
-        """

gllm_inference/em_invoker/google_em_invoker.pyi

@@ -5,7 +5,7 @@ from gllm_inference.em_invoker.em_invoker import BaseEMInvoker as BaseEMInvoker
 from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, Vector as Vector
 from typing import Any
 
-VALID_EXTENSIONS_MAP: Incomplete
+SUPPORTED_ATTACHMENTS: Incomplete
 
 class GoogleEMInvoker(BaseEMInvoker):
     '''An embedding model invoker to interact with Google embedding models.

gllm_inference/em_invoker/langchain/__init__.pyi

@@ -1,4 +1,3 @@
 from gllm_inference.em_invoker.langchain.em_invoker_embeddings import EMInvokerEmbeddings as EMInvokerEmbeddings
-from gllm_inference.em_invoker.langchain.tei_embeddings import TEIEmbeddings as TEIEmbeddings
 
-__all__ = ['EMInvokerEmbeddings', 'TEIEmbeddings']
+__all__ = ['EMInvokerEmbeddings']

gllm_inference/em_invoker/langchain_em_invoker.pyi

@@ -6,7 +6,7 @@ from gllm_inference.utils import load_langchain_model as load_langchain_model, p
 from langchain_core.embeddings import Embeddings as Embeddings
 from typing import Any
 
-VALID_EXTENSIONS_MAP: Incomplete
+SUPPORTED_ATTACHMENTS: Incomplete
 
 class LangChainEMInvoker(BaseEMInvoker):
     """A language model invoker to interact with LangChain's Embeddings.
@@ -19,7 +19,7 @@ class LangChainEMInvoker(BaseEMInvoker):
         retry_config (RetryConfig): The retry 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, em: Embeddings | 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) -> None:
         '''Initializes a new instance of the LangChainEMInvoker class.
 
         Args:
@@ -36,14 +36,4 @@ 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 is used.
-            em (Embeddings | None, optional): Deprecated parameter to pass the LangChain\'s Embeddings instance.
-                Equivalent to the `model` parameter. Retained for backward compatibility. Defaults to None.
         '''
-    def to_langchain(self) -> Embeddings:
-        """Converts the current embedding model invoker to an instance of LangChain `Embeddings` object.
-
-        This method converts the EM invoker to an instance of LangChain's `Embeddings` object.
-
-        Returns:
-            Embeddings: An instance of LangChain `Embeddings` object.
-        """

gllm_inference/em_invoker/openai_em_invoker.pyi

@@ -4,7 +4,7 @@ from gllm_inference.em_invoker.em_invoker import BaseEMInvoker as BaseEMInvoker
 from gllm_inference.schema import ModelId as ModelId, ModelProvider as ModelProvider, Vector as Vector
 from typing import Any
 
-VALID_EXTENSIONS_MAP: Incomplete
+SUPPORTED_ATTACHMENTS: Incomplete
 
 class OpenAIEMInvoker(BaseEMInvoker):
     '''An embedding model invoker to interact with OpenAI embedding models.

gllm_inference/em_invoker/twelevelabs_em_invoker.pyi

@@ -3,10 +3,9 @@ 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.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 langchain_core.embeddings import Embeddings as Embeddings
 from typing import Any
 
-VALID_EXTENSIONS_MAP: Incomplete
+SUPPORTED_ATTACHMENTS: Incomplete
 
 class TwelveLabsEMInvoker(BaseEMInvoker):
     '''An embedding model invoker to interact with TwelveLabs embedding models.
@@ -20,10 +19,7 @@ class TwelveLabsEMInvoker(BaseEMInvoker):
         retry_config (RetryConfig): The retry configuration for the embedding model.
 
     Input types:
-        The `TwelveLabsEMInvoker` supports the following input types:
-        1. Text.
-        2. Audio: ".mp3", ".wav", and ".flac".
-        3. Image: ".png", ".jpeg", and ".jpg".
+        The `TwelveLabsEMInvoker` supports the following input types: text, audio, and image.
         Non-text inputs must be passed as a `Attachment` object.
 
     Output format:
@@ -100,15 +96,3 @@ class TwelveLabsEMInvoker(BaseEMInvoker):
             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 is used.
         """
-    def to_langchain(self) -> Embeddings:
-        """Converts the current embedding model invoker to an instance of LangChain `Embeddings` object.
-
-        This method converts the EM invoker to an instance of LangChain's `Embeddings` object.
-        However, the TwelveLabsEMInvoker is not supported by LangChain.
-
-        Returns:
-            Embeddings: An instance of LangChain `Embeddings` object.
-
-        Raises:
-            NotImplementedError: The TwelveLabsEMInvoker is not supported by LangChain.
-        """

gllm_inference/em_invoker/voyage_em_invoker.pyi

@@ -5,7 +5,7 @@ from gllm_inference.em_invoker.schema.voyage import InputType as InputType, 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 typing import Any
 
-VALID_EXTENSIONS_MAP: Incomplete
+SUPPORTED_ATTACHMENTS: Incomplete
 MAX_PYTHON_MINOR_VERSION: int
 
 class VoyageEMInvoker(BaseEMInvoker):
@@ -20,10 +20,7 @@ class VoyageEMInvoker(BaseEMInvoker):
         retry_config (RetryConfig): The retry configuration for the embedding model.
 
     Input types:
-        The `VoyageEMInvoker` supports the following input types:
-        1. Text.
-        2. Image: ".png", ".jpeg", and ".jpg".
-        3. A tuple containing text and image.
+        The `VoyageEMInvoker` supports the following input types: text, image, and a tuple containing text and image.
         Non-text inputs must be passed as a `Attachment` object.
 
     Output format:

gllm_inference/lm_invoker/__init__.pyi

@@ -2,13 +2,10 @@ from gllm_inference.lm_invoker.anthropic_lm_invoker import AnthropicLMInvoker as
 from gllm_inference.lm_invoker.azure_openai_lm_invoker import AzureOpenAILMInvoker as AzureOpenAILMInvoker
 from gllm_inference.lm_invoker.bedrock_lm_invoker import BedrockLMInvoker as BedrockLMInvoker
 from gllm_inference.lm_invoker.datasaur_lm_invoker import DatasaurLMInvoker as DatasaurLMInvoker
-from gllm_inference.lm_invoker.google_generativeai_lm_invoker import GoogleGenerativeAILMInvoker as GoogleGenerativeAILMInvoker
 from gllm_inference.lm_invoker.google_lm_invoker import GoogleLMInvoker as GoogleLMInvoker
-from gllm_inference.lm_invoker.google_vertexai_lm_invoker import GoogleVertexAILMInvoker as GoogleVertexAILMInvoker
 from gllm_inference.lm_invoker.langchain_lm_invoker import LangChainLMInvoker as LangChainLMInvoker
 from gllm_inference.lm_invoker.litellm_lm_invoker import LiteLLMLMInvoker as LiteLLMLMInvoker
 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.tgi_lm_invoker import TGILMInvoker as TGILMInvoker
 
-__all__ = ['AnthropicLMInvoker', 'AzureOpenAILMInvoker', 'BedrockLMInvoker', 'DatasaurLMInvoker', 'GoogleGenerativeAILMInvoker', 'GoogleLMInvoker', 'GoogleVertexAILMInvoker', 'LangChainLMInvoker', 'LiteLLMLMInvoker', 'OpenAICompatibleLMInvoker', 'OpenAILMInvoker', 'TGILMInvoker']
+__all__ = ['AnthropicLMInvoker', 'AzureOpenAILMInvoker', 'BedrockLMInvoker', 'DatasaurLMInvoker', 'GoogleLMInvoker', 'LangChainLMInvoker', 'LiteLLMLMInvoker', 'OpenAICompatibleLMInvoker', 'OpenAILMInvoker']

gllm_inference/lm_invoker/anthropic_lm_invoker.pyi

@@ -3,11 +3,11 @@ from gllm_core.event import EventEmitter as EventEmitter
 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.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, ModelId as ModelId, ModelProvider as ModelProvider, MultimodalPrompt as MultimodalPrompt, Reasoning as Reasoning, ResponseSchema as ResponseSchema, TokenUsage as TokenUsage, ToolCall as ToolCall, ToolResult as ToolResult
+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 typing import Any
 
-VALID_EXTENSIONS_MAP: Incomplete
+SUPPORTED_ATTACHMENTS: Incomplete
 DEFAULT_MAX_TOKENS: int
 DEFAULT_THINKING_BUDGET: int
 
@@ -27,8 +27,6 @@ class AnthropicLMInvoker(BaseLMInvoker):
         retry_config (RetryConfig): The retry configuration for the language model.
         thinking (bool): Whether to enable thinking. Only allowed for thinking models.
         thinking_budget (int): The tokens allocated for the thinking process. Only allowed for thinking models.
-        output_thinking (bool): Whether to output the thinking token. Will be removed in v0.5.0, where the thinking
-            token will always be included in the output when thinking is enabled.
 
     Basic usage:
         The `AnthropicLMInvoker` can be used as follows:
@@ -38,21 +36,14 @@ class AnthropicLMInvoker(BaseLMInvoker):
         ```
 
     Input types:
-        The `AnthropicLMInvoker` supports the following input types:
-        1. Text.
-        2. Document: ".pdf".
-        3. Image: ".jpg", ".jpeg", ".png", ".gif", and ".webp".
-        Non-text inputs must be of valid file extensions and can be passed as an `Attachment` object.
-
-        Non-text inputs can only be passed with the `user` role.
+        The `AnthropicLMInvoker` supports the following input types: text, image, and document.
+        Non-text inputs can be passed as an `Attachment` object with the `user` role.
 
         Usage example:
         ```python
         text = "What animal is in this image?"
         image = Attachment.from_path("path/to/local/image.png")
-
-        prompt = [(PromptRole.USER, [text, image])]
-        result = await lm_invoker.invoke(prompt)
+        result = await lm_invoker.invoke([text, image])
         ```
 
     Tool calling:
@@ -180,8 +171,6 @@ class AnthropicLMInvoker(BaseLMInvoker):
            allocate at most 1024 tokens for thinking and the remaining 1024 tokens for generating the response.
 
         When enabled, the reasoning is stored in the `reasoning` attribute in the output.
-        **Note**: Before v0.5.0, this has to be set explicitly via the `output_thinking` parameter.
-        Starting from v0.5.0, the reasoning will always be included in the output when thinking is enabled.
 
         Usage example:
         ```python
@@ -209,7 +198,7 @@ class AnthropicLMInvoker(BaseLMInvoker):
         ```
 
     Output types:
-        The output of the `AnthropicLMInvoker` is of type `MultimodalOutput`, which is a type alias that can represent:
+        The output of the `AnthropicLMInvoker` can either be:
         1. `str`: The text response if no additional output is needed.
         2. `LMOutput`: A Pydantic model with the following attributes if any additional output is needed:
             2.1. response (str): The text response.
@@ -232,8 +221,7 @@ class AnthropicLMInvoker(BaseLMInvoker):
     client: Incomplete
     thinking: Incomplete
     thinking_budget: Incomplete
-    output_thinking: 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 = ..., use_thinking: bool = False, output_thinking: bool = False, bind_tools_params: dict[str, Any] | None = None, with_structured_output_params: dict[str, Any] | 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, 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:
         """Initializes the AnthropicLmInvoker instance.
 
         Args:
@@ -253,20 +241,10 @@ class AnthropicLMInvoker(BaseLMInvoker):
             thinking (bool, optional): Whether to enable thinking. Only allowed for thinking models. Defaults to False.
             thinking_budget (int, optional): The tokens allocated for the thinking process. Must be greater than or
                 equal to 1024. Only allowed for thinking models. Defaults to DEFAULT_THINKING_BUDGET.
-            use_thinking (bool, optional): Deprecated parameter to enable thinking. Defaults to False.
-            output_thinking (bool, optional): Deprecated parameter to output the thinking token. Starting from v0.5.0,
-                the thinking token will always be included in the output when thinking is enabled. Defaults to False.
-            bind_tools_params (dict[str, Any] | None, optional): Deprecated parameter to add tool calling capability.
-                If provided, must at least include the `tools` key that is equivalent to the `tools` parameter.
-                Retained for backward compatibility. Defaults to None.
-            with_structured_output_params (dict[str, Any] | None, optional): Deprecated parameter to instruct the
-                model to produce output with a certain schema. If provided, must at least include the `schema` key that
-                is equivalent to the `response_schema` parameter. Retained for backward compatibility. Defaults to None.
 
         Raises:
             ValueError:
             1. `thinking` is True, but the `thinking_budget` is less than 1024.
-            2. `thinking` is True and `tools` are provided, but `output_thinking` is False.  # TODO: Remove in v0.5.0
             3. `response_schema` is provided, but `tools` or `thinking` are also provided.
         """
     def set_tools(self, tools: list[Tool]) -> None:

gllm_inference/lm_invoker/azure_openai_lm_invoker.pyi

@@ -38,20 +38,14 @@ class AzureOpenAILMInvoker(OpenAILMInvoker):
         ```
 
     Input types:
-        1. Text.
-        2. Document: ".pdf".
-        3. Image: ".jpg", ".jpeg", ".png", ".gif", and ".webp".
-        Non-text inputs must be of valid file extensions and can be passed as an `Attachment` object.
-
-        Non-text inputs can only be passed with the `user` role.
+        The `AzureOpenAILMInvoker` supports the following input types: text, document, and image.
+        Non-text inputs can be passed as an `Attachment` object with the `user` role.
 
         Usage example:
         ```python
         text = "What animal is in this image?"
         image = Attachment.from_path("path/to/local/image.png")
-
-        prompt = [(PromptRole.USER, [text, image])]
-        result = await lm_invoker.invoke(prompt)
+        result = await lm_invoker.invoke([text, image])
         ```
 
     Tool calling:
@@ -203,8 +197,7 @@ class AzureOpenAILMInvoker(OpenAILMInvoker):
         Setting reasoning-related parameters for non-reasoning models will raise an error.
 
     Output types:
-        The output of the `AzureOpenAILMInvoker` is of type `MultimodalOutput`, which is a type alias that can
-        represent:
+        The output of the `AzureOpenAILMInvoker` can either be:
         1. `str`: The text response if no additional output is needed.
         2. `LMOutput`: A Pydantic model with the following attributes if any additional output is needed:
             2.1. response (str): The text response.
@@ -225,7 +218,7 @@ 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, bind_tools_params: dict[str, Any] | None = None, with_structured_output_params: dict[str, Any] | 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] | 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:
@@ -250,12 +243,6 @@ class AzureOpenAILMInvoker(OpenAILMInvoker):
                 for non-reasoning models. If None, the model will perform medium reasoning effort. Defaults to None.
             reasoning_summary (ReasoningSummary | None, optional): The reasoning summary level for reasoning models.
                 Not allowed for non-reasoning models. If None, no summary will be generated. Defaults to None.
-            bind_tools_params (dict[str, Any] | None, optional): Deprecated parameter to add tool calling capability.
-                If provided, must at least include the `tools` key that is equivalent to the `tools` parameter.
-                Retained for backward compatibility. Defaults to None.
-            with_structured_output_params (dict[str, Any] | None, optional): Deprecated parameter to instruct the
-                model to produce output with a certain schema. If provided, must at least include the `schema` key that
-                is equivalent to the `response_schema` parameter. Retained for backward compatibility. Defaults to None.
 
         Raises:
             ValueError: