langchain-google-genai 2.1.2__tar.gz → 2.1.3__tar.gz

{langchain_google_genai-2.1.2 → langchain_google_genai-2.1.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: langchain-google-genai
-Version: 2.1.2
+Version: 2.1.3
 Summary: An integration package connecting Google's genai package and LangChain
 Home-page: https://github.com/langchain-ai/langchain-google
 License: MIT
@@ -13,7 +13,7 @@ Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Requires-Dist: filetype (>=1.2.0,<2.0.0)
 Requires-Dist: google-ai-generativelanguage (>=0.6.16,<0.7.0)
-Requires-Dist: langchain-core (>=0.3.49,<0.4.0)
+Requires-Dist: langchain-core (>=0.3.52,<0.4.0)
 Requires-Dist: pydantic (>=2,<3)
 Project-URL: Repository, https://github.com/langchain-ai/langchain-google
 Project-URL: Source Code, https://github.com/langchain-ai/langchain-google/tree/main/libs/genai

{langchain_google_genai-2.1.2 → langchain_google_genai-2.1.3}/langchain_google_genai/_common.py

@@ -36,7 +36,7 @@ Supported examples:
     "the GOOGLE_API_KEY envvar"
     temperature: float = 0.7
     """Run inference with this temperature. Must by in the closed interval
-       [0.0, 1.0]."""
+       [0.0, 2.0]."""
     top_p: Optional[float] = None
     """Decode using nucleus sampling: consider the smallest set of tokens whose
        probability sum is at least top_p. Must be in the closed interval [0.0, 1.0]."""

{langchain_google_genai-2.1.2 → langchain_google_genai-2.1.3}/langchain_google_genai/chat_models.py

@@ -1,10 +1,13 @@
 from __future__ import annotations
 
 import asyncio
+import base64
 import json
 import logging
+import mimetypes
 import uuid
 import warnings
+from difflib import get_close_matches
 from operator import itemgetter
 from typing import (
     Any,
@@ -22,6 +25,7 @@ from typing import (
     cast,
 )
 
+import filetype  # type: ignore[import]
 import google.api_core
 
 # TODO: remove ignore once the google package is published with types
@@ -61,6 +65,7 @@ from langchain_core.messages import (
     HumanMessage,
     SystemMessage,
     ToolMessage,
+    is_data_content_block,
 )
 from langchain_core.messages.ai import UsageMetadata
 from langchain_core.messages.tool import invalid_tool_call, tool_call, tool_call_chunk
@@ -112,9 +117,6 @@ from langchain_google_genai._image_utils import (
 
 from . import _genai_extension as genaix
 
-WARNED_STRUCTURED_OUTPUT_JSON_MODE = False
-
-
 logger = logging.getLogger(__name__)
 
 
@@ -239,7 +241,7 @@ async def _achat_with_retry(generation_method: Callable, **kwargs: Any) -> Any:
     return await _achat_with_retry(**kwargs)
 
 
-def _is_openai_parts_format(part: dict) -> bool:
+def _is_lc_content_block(part: dict) -> bool:
     return "type" in part
 
 
@@ -254,10 +256,29 @@ def _convert_to_parts(
         if isinstance(part, str):
             parts.append(Part(text=part))
         elif isinstance(part, Mapping):
-            # OpenAI Format
-            if _is_openai_parts_format(part):
+            if _is_lc_content_block(part):
                 if part["type"] == "text":
                     parts.append(Part(text=part["text"]))
+                elif is_data_content_block(part):
+                    if part["source_type"] == "url":
+                        bytes_ = image_loader._bytes_from_url(part["url"])
+                    elif part["source_type"] == "base64":
+                        bytes_ = base64.b64decode(part["data"])
+                    else:
+                        raise ValueError("source_type must be url or base64.")
+                    inline_data: dict = {"data": bytes_}
+                    if "mime_type" in part:
+                        inline_data["mime_type"] = part["mime_type"]
+                    else:
+                        source = cast(str, part.get("url") or part.get("data"))
+                        mime_type, _ = mimetypes.guess_type(source)
+                        if not mime_type:
+                            kind = filetype.guess(bytes_)
+                            if kind:
+                                mime_type = kind.mime
+                        if mime_type:
+                            inline_data["mime_type"] = mime_type
+                    parts.append(Part(inline_data=inline_data))
                 elif part["type"] == "image_url":
                     img_url = part["image_url"]
                     if isinstance(img_url, dict):
@@ -709,7 +730,7 @@ class ChatGoogleGenerativeAI(_BaseGoogleGenerativeAI, BaseChatModel):
         .. code-block:: python
 
             AIMessageChunk(content='J', response_metadata={'finish_reason': 'STOP', 'safety_ratings': []}, id='run-e905f4f4-58cb-4a10-a960-448a2bb649e3', usage_metadata={'input_tokens': 18, 'output_tokens': 1, 'total_tokens': 19})
-            AIMessageChunk(content="'adore programmer. \n", response_metadata={'finish_reason': 'STOP', 'safety_ratings': [{'category': 'HARM_CATEGORY_SEXUALLY_EXPLICIT', 'probability': 'NEGLIGIBLE', 'blocked': False}, {'category': 'HARM_CATEGORY_HATE_SPEECH', 'probability': 'NEGLIGIBLE', 'blocked': False}, {'category': 'HARM_CATEGORY_HARASSMENT', 'probability': 'NEGLIGIBLE', 'blocked': False}, {'category': 'HARM_CATEGORY_DANGEROUS_CONTENT', 'probability': 'NEGLIGIBLE', 'blocked': False}]}, id='run-e905f4f4-58cb-4a10-a960-448a2bb649e3', usage_metadata={'input_tokens': 18, 'output_tokens': 5, 'total_tokens': 23})
+            AIMessageChunk(content="'adore programmer. \\n", response_metadata={'finish_reason': 'STOP', 'safety_ratings': [{'category': 'HARM_CATEGORY_SEXUALLY_EXPLICIT', 'probability': 'NEGLIGIBLE', 'blocked': False}, {'category': 'HARM_CATEGORY_HATE_SPEECH', 'probability': 'NEGLIGIBLE', 'blocked': False}, {'category': 'HARM_CATEGORY_HARASSMENT', 'probability': 'NEGLIGIBLE', 'blocked': False}, {'category': 'HARM_CATEGORY_DANGEROUS_CONTENT', 'probability': 'NEGLIGIBLE', 'blocked': False}]}, id='run-e905f4f4-58cb-4a10-a960-448a2bb649e3', usage_metadata={'input_tokens': 18, 'output_tokens': 5, 'total_tokens': 23})
 
         .. code-block:: python
 
@@ -739,6 +760,109 @@ class ChatGoogleGenerativeAI(_BaseGoogleGenerativeAI, BaseChatModel):
             # batch:
             # await llm.abatch([messages])
 
+    Context Caching:
+        Context caching allows you to store and reuse content (e.g., PDFs, images) for faster processing.
+        The `cached_content` parameter accepts a cache name created via the Google Generative AI API.
+        Below are two examples: caching a single file directly and caching multiple files using `Part`.
+
+        Single File Example:
+        This caches a single file and queries it.
+
+        .. code-block:: python
+
+            from google import genai
+            from google.genai import types
+            import time
+            from langchain_google_genai import ChatGoogleGenerativeAI
+            from langchain_core.messages import HumanMessage
+
+            client = genai.Client()
+
+            # Upload file
+            file = client.files.upload(file="./example_file")
+            while file.state.name == 'PROCESSING':
+                time.sleep(2)
+                file = client.files.get(name=file.name)
+
+            # Create cache
+            model = 'models/gemini-1.5-flash-001'
+            cache = client.caches.create(
+                model=model,
+                config=types.CreateCachedContentConfig(
+                    display_name='Cached Content',
+                    system_instruction=(
+                        'You are an expert content analyzer, and your job is to answer '
+                        'the user\'s query based on the file you have access to.'
+                    ),
+                    contents=[file],
+                    ttl="300s",
+                )
+            )
+
+            # Query with LangChain
+            llm = ChatGoogleGenerativeAI(
+                model=model,
+                cached_content=cache.name,
+            )
+            message = HumanMessage(content="Summarize the main points of the content.")
+            llm.invoke([message])
+
+        Multiple Files Example:
+        This caches two files using `Part` and queries them together.
+
+        .. code-block:: python
+
+            from google import genai
+            from google.genai.types import CreateCachedContentConfig, Content, Part
+            import time
+            from langchain_google_genai import ChatGoogleGenerativeAI
+            from langchain_core.messages import HumanMessage
+
+            client = genai.Client()
+
+            # Upload files
+            file_1 = client.files.upload(file="./file1")
+            while file_1.state.name == 'PROCESSING':
+                time.sleep(2)
+                file_1 = client.files.get(name=file_1.name)
+
+            file_2 = client.files.upload(file="./file2")
+            while file_2.state.name == 'PROCESSING':
+                time.sleep(2)
+                file_2 = client.files.get(name=file_2.name)
+
+            # Create cache with multiple files
+            contents = [
+                Content(
+                    role="user",
+                    parts=[
+                        Part.from_uri(file_uri=file_1.uri, mime_type=file_1.mime_type),
+                        Part.from_uri(file_uri=file_2.uri, mime_type=file_2.mime_type),
+                    ],
+                )
+            ]
+            model = "gemini-1.5-flash-001"
+            cache = client.caches.create(
+                model=model,
+                config=CreateCachedContentConfig(
+                    display_name='Cached Contents',
+                    system_instruction=(
+                        'You are an expert content analyzer, and your job is to answer '
+                        'the user\'s query based on the files you have access to.'
+                    ),
+                    contents=contents,
+                    ttl="300s",
+                )
+            )
+
+            # Query with LangChain
+            llm = ChatGoogleGenerativeAI(
+                model=model,
+                cached_content=cache.name,
+            )
+            message = HumanMessage(content="Provide a summary of the key information across both files.")
+            llm.invoke([message])
+
     Tool calling:
         .. code-block:: python
 
@@ -842,7 +966,7 @@ class ChatGoogleGenerativeAI(_BaseGoogleGenerativeAI, BaseChatModel):
 
         .. code-block:: python
 
-            'The weather in this image appears to be sunny and pleasant. The sky is a bright blue with scattered white clouds, suggesting fair weather. The lush green grass and trees indicate a warm and possibly slightly breezy day. There are no signs of rain or storms. \n'
+            'The weather in this image appears to be sunny and pleasant. The sky is a bright blue with scattered white clouds, suggesting fair weather. The lush green grass and trees indicate a warm and possibly slightly breezy day. There are no signs of rain or storms.'
 
     Token usage:
         .. code-block:: python
@@ -891,6 +1015,28 @@ class ChatGoogleGenerativeAI(_BaseGoogleGenerativeAI, BaseChatModel):
     ``cachedContents/{cachedContent}``.
     """
 
+    def __init__(self, **kwargs: Any) -> None:
+        """Needed for arg validation."""
+        # Get all valid field names, including aliases
+        valid_fields = set()
+        for field_name, field_info in self.model_fields.items():
+            valid_fields.add(field_name)
+            if hasattr(field_info, "alias") and field_info.alias is not None:
+                valid_fields.add(field_info.alias)
+
+        # Check for unrecognized arguments
+        for arg in kwargs:
+            if arg not in valid_fields:
+                suggestions = get_close_matches(arg, valid_fields, n=1)
+                suggestion = (
+                    f" Did you mean: '{suggestions[0]}'?" if suggestions else ""
+                )
+                logger.warning(
+                    f"Unexpected argument '{arg}' "
+                    f"provided to ChatGoogleGenerativeAI.{suggestion}"
+                )
+        super().__init__(**kwargs)
+
     model_config = ConfigDict(
         populate_by_name=True,
     )
@@ -927,7 +1073,9 @@ class ChatGoogleGenerativeAI(_BaseGoogleGenerativeAI, BaseChatModel):
         if self.top_k is not None and self.top_k <= 0:
             raise ValueError("top_k must be positive")
 
-        if not self.model.startswith("models/"):
+        if not any(
+            self.model.startswith(prefix) for prefix in ("models/", "tunedModels/")
+        ):
             self.model = f"models/{self.model}"
 
         additional_headers = self.additional_headers or {}
@@ -964,12 +1112,17 @@ class ChatGoogleGenerativeAI(_BaseGoogleGenerativeAI, BaseChatModel):
         # this check ensures that async client is only initialized
         # within an asyncio event loop to avoid the error
         if not self.async_client_running and _is_event_loop_running():
+            # async clients don't support "rest" transport
+            # https://github.com/googleapis/gapic-generator-python/issues/1962
+            transport = self.transport
+            if transport == "rest":
+                transport = "grpc_asyncio"
             self.async_client_running = genaix.build_generative_async_service(
                 credentials=self.credentials,
                 api_key=google_api_key,
                 client_info=get_client_info("ChatGoogleGenerativeAI"),
                 client_options=self.client_options,
-                transport=self.transport,
+                transport=transport,
             )
         return self.async_client_running
 
@@ -1406,14 +1559,6 @@ class ChatGoogleGenerativeAI(_BaseGoogleGenerativeAI, BaseChatModel):
                 tools=[schema], first_tool_only=True
             )
         else:
-            global WARNED_STRUCTURED_OUTPUT_JSON_MODE
-            warnings.warn(
-                "ChatGoogleGenerativeAI.with_structured_output with dict schema has "
-                "changed recently to align with behavior of other LangChain chat "
-                "models. More context: "
-                "https://github.com/langchain-ai/langchain-google/pull/772"
-            )
-            WARNED_STRUCTURED_OUTPUT_JSON_MODE = True
             parser = JsonOutputKeyToolsParser(key_name=tool_name, first_tool_only=True)
         tool_choice = tool_name if self._supports_tool_choice else None
         try:

{langchain_google_genai-2.1.2 → langchain_google_genai-2.1.3}/langchain_google_genai/embeddings.py

@@ -6,6 +6,7 @@ from typing import Any, Dict, List, Optional
 from google.ai.generativelanguage_v1beta.types import (
     BatchEmbedContentsRequest,
     EmbedContentRequest,
+    EmbedContentResponse,
 )
 from langchain_core.embeddings import Embeddings
 from langchain_core.utils import secret_from_env
@@ -239,7 +240,8 @@ class GoogleGenerativeAIEmbeddings(BaseModel, Embeddings):
         title: Optional[str] = None,
         output_dimensionality: Optional[int] = None,
     ) -> List[float]:
-        """Embed a text.
+        """Embed a text, using the non-batch endpoint:
+        https://ai.google.dev/api/rest/v1/models/embedContent#EmbedContentRequest
 
         Args:
             text: The text to embed.
@@ -247,15 +249,19 @@ class GoogleGenerativeAIEmbeddings(BaseModel, Embeddings):
             title: An optional title for the text.
             Only applicable when TaskType is RETRIEVAL_DOCUMENT.
             output_dimensionality: Optional reduced dimension for the output embedding.
-            https://ai.google.dev/api/rest/v1/models/batchEmbedContents#EmbedContentRequest
 
         Returns:
             Embedding for the text.
         """
         task_type = self.task_type or "RETRIEVAL_QUERY"
-        return self.embed_documents(
-            [text],
-            task_type=task_type,
-            titles=[title] if title else None,
-            output_dimensionality=output_dimensionality,
-        )[0]
+        try:
+            request: EmbedContentRequest = self._prepare_request(
+                text=text,
+                task_type=task_type,
+                title=title,
+                output_dimensionality=output_dimensionality,
+            )
+            result: EmbedContentResponse = self.client.embed_content(request)
+        except Exception as e:
+            raise GoogleGenerativeAIError(f"Error embedding content: {e}") from e
+        return list(result.embedding.values)

{langchain_google_genai-2.1.2 → langchain_google_genai-2.1.3}/langchain_google_genai/llms.py

@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+import logging
+from difflib import get_close_matches
 from typing import Any, Iterator, List, Optional
 
 from langchain_core.callbacks import (
@@ -17,6 +19,8 @@ from langchain_google_genai._common import (
 )
 from langchain_google_genai.chat_models import ChatGoogleGenerativeAI
 
+logger = logging.getLogger(__name__)
+
 
 class GoogleGenerativeAI(_BaseGoogleGenerativeAI, BaseLLM):
     """Google GenerativeAI models.
@@ -33,6 +37,28 @@ class GoogleGenerativeAI(_BaseGoogleGenerativeAI, BaseLLM):
         populate_by_name=True,
     )
 
+    def __init__(self, **kwargs: Any) -> None:
+        """Needed for arg validation."""
+        # Get all valid field names, including aliases
+        valid_fields = set()
+        for field_name, field_info in self.model_fields.items():
+            valid_fields.add(field_name)
+            if hasattr(field_info, "alias") and field_info.alias is not None:
+                valid_fields.add(field_info.alias)
+
+        # Check for unrecognized arguments
+        for arg in kwargs:
+            if arg not in valid_fields:
+                suggestions = get_close_matches(arg, valid_fields, n=1)
+                suggestion = (
+                    f" Did you mean: '{suggestions[0]}'?" if suggestions else ""
+                )
+                logger.warning(
+                    f"Unexpected argument '{arg}' "
+                    f"provided to GoogleGenerativeAI.{suggestion}"
+                )
+        super().__init__(**kwargs)
+
     @model_validator(mode="after")
     def validate_environment(self) -> Self:
         """Validates params and passes them to google-generativeai package."""

{langchain_google_genai-2.1.2 → langchain_google_genai-2.1.3}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "langchain-google-genai"
-version = "2.1.2"
+version = "2.1.3"
 description = "An integration package connecting Google's genai package and LangChain"
 authors = []
 readme = "README.md"
@@ -12,7 +12,7 @@ license = "MIT"
 
 [tool.poetry.dependencies]
 python = ">=3.9,<4.0"
-langchain-core = "^0.3.49"
+langchain-core = "^0.3.52"
 google-ai-generativelanguage = "^0.6.16"
 pydantic = ">=2,<3"
 filetype = "^1.2.0"
@@ -27,8 +27,8 @@ pytest-mock = "^3.10.0"
 syrupy = "^4.0.2"
 pytest-watcher = "^0.3.4"
 pytest-asyncio = "^0.21.1"
-numpy = "^1.26.2"
-langchain-tests = "0.3.17"
+numpy = ">=1.26.2"
+langchain-tests = "0.3.18"
 
 [tool.codespell]
 ignore-words-list = "rouge"
@@ -60,7 +60,7 @@ mypy = "^1.10"
 types-requests = "^2.28.11.5"
 types-google-cloud-ndb = "^2.2.0.1"
 types-protobuf = "^4.24.0.20240302"
-numpy = "^1.26.2"
+numpy = ">=1.26.2"
 
 
 [tool.poetry.group.dev]