From 4af3f89a3a122080506a9d9ee1af8b9fde2f358b Mon Sep 17 00:00:00 2001 From: Sydney Runkle <54324534+sydney-runkle@users.noreply.github.com> Date: Wed, 16 Apr 2025 08:45:40 -0400 Subject: [PATCH 1/7] docs: enforce newlines when signature exceeds char threshold (#30866) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Below is an example of the single line vs new multiline approach. Before this PR: Screenshot 2025-04-15 at 8 56 26 PM After this PR: Screenshot 2025-04-15 at 8 56 13 PM --- docs/api_reference/conf.py | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/docs/api_reference/conf.py b/docs/api_reference/conf.py index b6a0464cb8f..514a17ad9b0 100644 --- a/docs/api_reference/conf.py +++ b/docs/api_reference/conf.py @@ -275,3 +275,7 @@ if os.environ.get("READTHEDOCS", "") == "True": html_context["READTHEDOCS"] = True master_doc = "index" + +# If a signature’s length in characters exceeds 60, +# each parameter within the signature will be displayed on an individual logical line +maximum_signature_line_length = 60 From ca39680d2ab0d786bc035930778a5787e7bb5e01 Mon Sep 17 00:00:00 2001 From: ccurme Date: Wed, 16 Apr 2025 09:14:57 -0400 Subject: [PATCH 2/7] ollama: release 0.3.2 (#30865) --- libs/partners/ollama/pyproject.toml | 4 ++-- libs/partners/ollama/uv.lock | 24 ++++++++++++++---------- 2 files changed, 16 insertions(+), 12 deletions(-) diff --git a/libs/partners/ollama/pyproject.toml b/libs/partners/ollama/pyproject.toml index 6d9a54af8ad..99d3a4a1f1e 100644 --- a/libs/partners/ollama/pyproject.toml +++ b/libs/partners/ollama/pyproject.toml @@ -6,9 +6,9 @@ build-backend = "pdm.backend" authors = [] license = { text = "MIT" } requires-python = "<4.0,>=3.9" -dependencies = ["ollama<1,>=0.4.4", "langchain-core<1.0.0,>=0.3.51"] +dependencies = ["ollama<1,>=0.4.4", "langchain-core<1.0.0,>=0.3.52"] name = "langchain-ollama" -version = "0.3.1" +version = "0.3.2" description = "An integration package connecting Ollama and LangChain" readme = "README.md" diff --git a/libs/partners/ollama/uv.lock b/libs/partners/ollama/uv.lock index 34d8f6ca1fd..895083899a4 100644 --- a/libs/partners/ollama/uv.lock +++ b/libs/partners/ollama/uv.lock @@ -1,7 +1,8 @@ version = 1 requires-python = ">=3.9, <4.0" resolution-markers = [ - "python_full_version >= '3.12.4'", + "python_full_version >= '3.13'", + "python_full_version >= '3.12.4' and python_full_version < '3.13'", "python_full_version >= '3.12' and python_full_version < '3.12.4'", "python_full_version < '3.12'", ] @@ -287,7 +288,7 @@ wheels = [ [[package]] name = "langchain-core" -version = "0.3.51" +version = "0.3.52" source = { editable = "../../core" } dependencies = [ { name = "jsonpatch" }, @@ -323,10 +324,12 @@ test = [ { name = "freezegun", specifier = ">=1.2.2,<2.0.0" }, { name = "grandalf", specifier = ">=0.8,<1.0" }, { name = "langchain-tests", directory = "../../standard-tests" }, - { name = "numpy", marker = "python_full_version < '3.12'", specifier = ">=1.24.0,<2.0.0" }, - { name = "numpy", marker = "python_full_version >= '3.12'", specifier = ">=1.26.0,<3" }, + { name = "numpy", marker = "python_full_version < '3.13'", specifier = ">=1.26.4" }, + { name = "numpy", marker = "python_full_version >= '3.13'", specifier = ">=2.1.0" }, { name = "pytest", specifier = ">=8,<9" }, { name = "pytest-asyncio", specifier = ">=0.21.1,<1.0.0" }, + { name = "pytest-benchmark" }, + { name = "pytest-codspeed" }, { name = "pytest-mock", specifier = ">=3.10.0,<4.0.0" }, { name = "pytest-socket", specifier = ">=0.7.0,<1.0.0" }, { name = "pytest-watcher", specifier = ">=0.3.4,<1.0.0" }, @@ -337,15 +340,14 @@ test = [ test-integration = [] typing = [ { name = "langchain-text-splitters", directory = "../../text-splitters" }, - { name = "mypy", specifier = ">=1.10,<1.11" }, - { name = "types-jinja2", specifier = ">=2.11.9,<3.0.0" }, + { name = "mypy", specifier = ">=1.15,<1.16" }, { name = "types-pyyaml", specifier = ">=6.0.12.2,<7.0.0.0" }, { name = "types-requests", specifier = ">=2.28.11.5,<3.0.0.0" }, ] [[package]] name = "langchain-ollama" -version = "0.3.1" +version = "0.3.2" source = { editable = "." } dependencies = [ { name = "langchain-core" }, @@ -403,7 +405,7 @@ typing = [ [[package]] name = "langchain-tests" -version = "0.3.17" +version = "0.3.18" source = { editable = "../../standard-tests" } dependencies = [ { name = "httpx" }, @@ -420,7 +422,8 @@ dependencies = [ requires-dist = [ { name = "httpx", specifier = ">=0.25.0,<1" }, { name = "langchain-core", editable = "../../core" }, - { name = "numpy", specifier = ">=1.26.2,<3" }, + { name = "numpy", marker = "python_full_version < '3.13'", specifier = ">=1.26.2" }, + { name = "numpy", marker = "python_full_version >= '3.13'", specifier = ">=2.1.0" }, { name = "pytest", specifier = ">=7,<9" }, { name = "pytest-asyncio", specifier = ">=0.20,<1" }, { name = "pytest-socket", specifier = ">=0.6.0,<1" }, @@ -558,7 +561,8 @@ name = "numpy" version = "2.2.2" source = { registry = "https://pypi.org/simple" } resolution-markers = [ - "python_full_version >= '3.12.4'", + "python_full_version >= '3.13'", + "python_full_version >= '3.12.4' and python_full_version < '3.13'", "python_full_version >= '3.12' and python_full_version < '3.12.4'", ] sdist = { url = "https://files.pythonhosted.org/packages/ec/d0/c12ddfd3a02274be06ffc71f3efc6d0e457b0409c4481596881e748cb264/numpy-2.2.2.tar.gz", hash = "sha256:ed6906f61834d687738d25988ae117683705636936cc605be0bb208b23df4d8f", size = 20233295 } From 60d8ade078419339aec0074231c58946f26937f0 Mon Sep 17 00:00:00 2001 From: rrozanski-smabbler Date: Wed, 16 Apr 2025 17:39:04 +0300 Subject: [PATCH 3/7] Galaxia integration (#30792) - [ ] **PR title**: "docs: adding Smabbler's Galaxia integration" - [ ] **PR message**: **Twitter handle:** @Galaxia_graph I'm adding docs here + added the package to the packages.yml. I didn't add a unit test, because this integration is just a thin wrapper on top of our API. There isn't much left to test if you mock it away. --------- Co-authored-by: Chester Curme --- docs/docs/integrations/providers/galaxia.mdx | 32 +++ .../retrievers/galaxia-retriever.ipynb | 213 ++++++++++++++++++ libs/packages.yml | 4 + 3 files changed, 249 insertions(+) create mode 100644 docs/docs/integrations/providers/galaxia.mdx create mode 100644 docs/docs/integrations/retrievers/galaxia-retriever.ipynb diff --git a/docs/docs/integrations/providers/galaxia.mdx b/docs/docs/integrations/providers/galaxia.mdx new file mode 100644 index 00000000000..97855bfd7cf --- /dev/null +++ b/docs/docs/integrations/providers/galaxia.mdx @@ -0,0 +1,32 @@ +# Smabbler +> Smabbler’s graph-powered platform boosts AI development by transforming data into a structured knowledge foundation. + +# Galaxia + +> Galaxia Knowledge Base is an integrated knowledge base and retrieval mechanism for RAG. In contrast to standard solution, it is based on Knowledge Graphs built using symbolic NLP and Knowledge Representation solutions. Provided texts are analysed and transformed into Graphs containing text, language and semantic information. This rich structure allows for retrieval that is based on semantic information, not on vector similarity/distance. + +Implementing RAG using Galaxia involves first uploading your files to [Galaxia](https://beta.cloud.smabbler.com/home), analyzing them there and then building a model (knowledge graph). When the model is built, you can use `GalaxiaRetriever` to connect to the API and start retrieving. + +More information: [docs](https://smabbler.gitbook.io/smabbler) + +## Installation +``` +pip install langchain-galaxia-retriever +``` + +## Usage + +``` +from langchain_galaxia_retriever.retriever import GalaxiaRetriever + +gr = GalaxiaRetriever( + api_url="beta.api.smabbler.com", + api_key="", + knowledge_base_id="", + n_retries=10, + wait_time=5, +) + +result = gr.invoke('') +print(result) + diff --git a/docs/docs/integrations/retrievers/galaxia-retriever.ipynb b/docs/docs/integrations/retrievers/galaxia-retriever.ipynb new file mode 100644 index 00000000000..399b8b348d5 --- /dev/null +++ b/docs/docs/integrations/retrievers/galaxia-retriever.ipynb @@ -0,0 +1,213 @@ +{ + "cells": [ + { + "cell_type": "raw", + "id": "2af1fec5-4ca6-4167-8ee1-13314aac3258", + "metadata": { + "vscode": { + "languageId": "raw" + } + }, + "source": [ + "---\n", + "sidebar_label: Galaxia\n", + "---" + ] + }, + { + "cell_type": "markdown", + "id": "1d7d6cbc-4373-4fb5-94dd-acd610165452", + "metadata": {}, + "source": [ + "# Galaxia Retriever\n", + "\n", + "Galaxia is GraphRAG solution, which automates document processing, knowledge base (Graph Language Model) creation and retrieval:\n", + "[galaxia-rag](https://smabbler.gitbook.io/smabbler/api-rag/smabblers-api-rag)\n", + "\n", + "To use Galaxia first upload your texts and create a Graph Language Model here: [smabbler-cloud](https://beta.cloud.smabbler.com)\n", + "\n", + "After the model is built and activated, you will be able to use this integration to retrieve what you need.\n", + "\n", + "The module repository is located here: [github](https://github.com/rrozanski-smabbler/galaxia-langchain)\n", + "\n", + "### Integration details\n", + "| Retriever | Self-host | Cloud offering | Package |\n", + "| :--- | :--- | :---: | :---: |\n", + "[Galaxia Retriever](https://github.com/rrozanski-smabbler/galaxia-langchain) | ❌ | ✅ | __langchain-galaxia-retriever__ |" + ] + }, + { + "cell_type": "markdown", + "id": "82fa1c05-c205-4429-a74c-e6c81c4e8611", + "metadata": {}, + "source": [ + "## Setup\n", + "Before you can retrieve anything you need to create your Graph Language Model here: [smabbler-cloud](https://beta.cloud.smabbler.com)\n", + "\n", + "following these 3 simple steps: [rag-instruction](https://smabbler.gitbook.io/smabbler/api-rag/build-rag-model-in-3-steps)\n", + "\n", + "Don't forget to activate the model after building it!" + ] + }, + { + "cell_type": "markdown", + "id": "91897867-eb39-4c3b-8df8-5427043ecdcd", + "metadata": {}, + "source": [ + "### Installation\n", + "The retriever is implemented in the following package: [pypi](https://pypi.org/project/langchain-galaxia-retriever/)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ceca36f2-013c-4b28-81fe-8808d0cf6419", + "metadata": {}, + "outputs": [], + "source": [ + "%pip install -qU langchain-galaxia-retriever" + ] + }, + { + "cell_type": "markdown", + "id": "019e0e50-5e66-440b-9cf1-d21b4009bf13", + "metadata": {}, + "source": [ + "## Instantiation" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "c7188217-4b26-4201-b15a-b7a5f263f815", + "metadata": {}, + "outputs": [], + "source": [ + "from langchain_galaxia_retriever.retriever import GalaxiaRetriever\n", + "\n", + "gr = GalaxiaRetriever(\n", + " api_url=\"beta.api.smabbler.com\",\n", + " api_key=\"\", # you can find it here: https://beta.cloud.smabbler.com/user/account\n", + " knowledge_base_id=\"\", # you can find it in https://beta.cloud.smabbler.com , in the model table\n", + " n_retries=10,\n", + " wait_time=5,\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "02d288a5-4f76-472e-9a60-eea8e6b8dc7a", + "metadata": {}, + "source": [ + "## Usage" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "5f79e03f-77a6-4eb6-b41d-f3da2f897654", + "metadata": {}, + "outputs": [], + "source": [ + "result = gr.invoke(\"\")\n", + "print(result)" + ] + }, + { + "cell_type": "markdown", + "id": "ffb2a595-a901-477a-a374-efd091bc1c9a", + "metadata": {}, + "source": [ + "## Use within a chain" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "9c2e2394-ca33-47be-a851-551b4216daea", + "metadata": {}, + "outputs": [], + "source": [ + "# | output: false\n", + "# | echo: false\n", + "\n", + "from langchain_openai import ChatOpenAI\n", + "\n", + "llm = ChatOpenAI(model=\"gpt-3.5-turbo-0125\", temperature=0)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ed8699d6-d65d-40ea-8c58-8d809cc512cf", + "metadata": {}, + "outputs": [], + "source": [ + "from langchain_core.output_parsers import StrOutputParser\n", + "from langchain_core.prompts import ChatPromptTemplate\n", + "from langchain_core.runnables import RunnablePassthrough\n", + "\n", + "prompt = ChatPromptTemplate.from_template(\n", + " \"\"\"Answer the question based only on the context provided.\n", + "\n", + "Context: {context}\n", + "\n", + "Question: {question}\"\"\"\n", + ")\n", + "\n", + "\n", + "def format_docs(docs):\n", + " return \"\\n\\n\".join(doc.page_content for doc in docs)\n", + "\n", + "\n", + "chain = (\n", + " {\"context\": gr | format_docs, \"question\": RunnablePassthrough()}\n", + " | prompt\n", + " | llm\n", + " | StrOutputParser()\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f9b944d7-8800-4926-b1ce-fcdc52ecda1c", + "metadata": {}, + "outputs": [], + "source": [ + "chain.invoke(\"\")" + ] + }, + { + "cell_type": "markdown", + "id": "11b5c9a5-0a66-415f-98f8-f12080cad30a", + "metadata": {}, + "source": [ + "## API reference\n", + "\n", + "For more information about Galaxia Retriever check its implementation on github [github](https://github.com/rrozanski-smabbler/galaxia-langchain)" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3 (ipykernel)", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.11.7" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/libs/packages.yml b/libs/packages.yml index 6d2689e924a..cb8d0b43580 100644 --- a/libs/packages.yml +++ b/libs/packages.yml @@ -606,3 +606,7 @@ packages: - name: langchain-ydb path: . repo: ydb-platform/langchain-ydb +- name: langchain-galaxia-retriever + provider_page: galaxia + path: . + repo: rrozanski-smabbler/galaxia-langchain From 88fce67724123d4cabacffbc8a07ba97534a11ff Mon Sep 17 00:00:00 2001 From: Sydney Runkle <54324534+sydney-runkle@users.noreply.github.com> Date: Wed, 16 Apr 2025 12:00:08 -0400 Subject: [PATCH 4/7] core: Removing unnecessary `pydantic` core schema rebuilds (#30848) We only need to rebuild model schemas if type annotation information isn't available during declaration - that shouldn't be the case for these types corrected here. Need to do more thorough testing to make sure these structures have complete schemas, but hopefully this boosts startup / import time. --- libs/core/langchain_core/messages/ai.py | 3 --- libs/core/langchain_core/messages/chat.py | 3 --- libs/core/langchain_core/messages/function.py | 3 --- libs/core/langchain_core/messages/human.py | 3 --- libs/core/langchain_core/messages/modifier.py | 3 --- libs/core/langchain_core/messages/system.py | 3 --- libs/core/langchain_core/messages/tool.py | 3 --- .../langchain_core/output_parsers/list.py | 3 --- .../langchain_core/output_parsers/pydantic.py | 3 --- .../langchain_core/output_parsers/string.py | 3 --- libs/core/langchain_core/prompts/pipeline.py | 3 --- libs/core/langchain_core/runnables/base.py | 3 --- .../langchain_core/runnables/configurable.py | 4 ---- libs/core/langchain_core/runnables/utils.py | 3 +-- libs/core/langchain_core/tools/simple.py | 3 --- .../chat_models/test_rate_limiting.py | 3 --- .../output_parsers/test_base_parsers.py | 2 -- .../unit_tests/prompts/test_structured.py | 3 --- .../runnables/test_runnable_events_v1.py | 3 --- .../tests/unit_tests/test_pydantic_imports.py | 20 +++++++++++++++++++ libs/core/tests/unit_tests/test_tools.py | 3 --- uv.lock | 9 +++++---- 22 files changed, 26 insertions(+), 63 deletions(-) create mode 100644 libs/core/tests/unit_tests/test_pydantic_imports.py diff --git a/libs/core/langchain_core/messages/ai.py b/libs/core/langchain_core/messages/ai.py index 6c2a08e6db1..918f717aa58 100644 --- a/libs/core/langchain_core/messages/ai.py +++ b/libs/core/langchain_core/messages/ai.py @@ -276,9 +276,6 @@ class AIMessage(BaseMessage): return (base.strip() + "\n" + "\n".join(lines)).strip() -AIMessage.model_rebuild() - - class AIMessageChunk(AIMessage, BaseMessageChunk): """Message chunk from an AI.""" diff --git a/libs/core/langchain_core/messages/chat.py b/libs/core/langchain_core/messages/chat.py index c9f11f976b6..a4791423fad 100644 --- a/libs/core/langchain_core/messages/chat.py +++ b/libs/core/langchain_core/messages/chat.py @@ -22,9 +22,6 @@ class ChatMessage(BaseMessage): """The type of the message (used during serialization). Defaults to "chat".""" -ChatMessage.model_rebuild() - - class ChatMessageChunk(ChatMessage, BaseMessageChunk): """Chat Message chunk.""" diff --git a/libs/core/langchain_core/messages/function.py b/libs/core/langchain_core/messages/function.py index c79d54ae5b2..fc1018775b7 100644 --- a/libs/core/langchain_core/messages/function.py +++ b/libs/core/langchain_core/messages/function.py @@ -30,9 +30,6 @@ class FunctionMessage(BaseMessage): """The type of the message (used for serialization). Defaults to "function".""" -FunctionMessage.model_rebuild() - - class FunctionMessageChunk(FunctionMessage, BaseMessageChunk): """Function Message chunk.""" diff --git a/libs/core/langchain_core/messages/human.py b/libs/core/langchain_core/messages/human.py index 5baa3a38856..4e19904ab33 100644 --- a/libs/core/langchain_core/messages/human.py +++ b/libs/core/langchain_core/messages/human.py @@ -52,9 +52,6 @@ class HumanMessage(BaseMessage): super().__init__(content=content, **kwargs) -HumanMessage.model_rebuild() - - class HumanMessageChunk(HumanMessage, BaseMessageChunk): """Human Message chunk.""" diff --git a/libs/core/langchain_core/messages/modifier.py b/libs/core/langchain_core/messages/modifier.py index aa43b82adca..9cfc89c5760 100644 --- a/libs/core/langchain_core/messages/modifier.py +++ b/libs/core/langchain_core/messages/modifier.py @@ -26,6 +26,3 @@ class RemoveMessage(BaseMessage): raise ValueError(msg) super().__init__("", id=id, **kwargs) - - -RemoveMessage.model_rebuild() diff --git a/libs/core/langchain_core/messages/system.py b/libs/core/langchain_core/messages/system.py index f29f2c85b1d..d63bd53a0fe 100644 --- a/libs/core/langchain_core/messages/system.py +++ b/libs/core/langchain_core/messages/system.py @@ -46,9 +46,6 @@ class SystemMessage(BaseMessage): super().__init__(content=content, **kwargs) -SystemMessage.model_rebuild() - - class SystemMessageChunk(SystemMessage, BaseMessageChunk): """System Message chunk.""" diff --git a/libs/core/langchain_core/messages/tool.py b/libs/core/langchain_core/messages/tool.py index 3a1408809c9..42c9018e151 100644 --- a/libs/core/langchain_core/messages/tool.py +++ b/libs/core/langchain_core/messages/tool.py @@ -146,9 +146,6 @@ class ToolMessage(BaseMessage, ToolOutputMixin): super().__init__(content=content, **kwargs) -ToolMessage.model_rebuild() - - class ToolMessageChunk(ToolMessage, BaseMessageChunk): """Tool Message chunk.""" diff --git a/libs/core/langchain_core/output_parsers/list.py b/libs/core/langchain_core/output_parsers/list.py index 2f52564fcd4..3dba7e3af5c 100644 --- a/libs/core/langchain_core/output_parsers/list.py +++ b/libs/core/langchain_core/output_parsers/list.py @@ -133,9 +133,6 @@ class ListOutputParser(BaseTransformOutputParser[list[str]]): yield [part] -ListOutputParser.model_rebuild() - - class CommaSeparatedListOutputParser(ListOutputParser): """Parse the output of an LLM call to a comma-separated list.""" diff --git a/libs/core/langchain_core/output_parsers/pydantic.py b/libs/core/langchain_core/output_parsers/pydantic.py index f41a8fe181e..194cc3a2d36 100644 --- a/libs/core/langchain_core/output_parsers/pydantic.py +++ b/libs/core/langchain_core/output_parsers/pydantic.py @@ -114,9 +114,6 @@ class PydanticOutputParser(JsonOutputParser, Generic[TBaseModel]): return self.pydantic_object -PydanticOutputParser.model_rebuild() - - _PYDANTIC_FORMAT_INSTRUCTIONS = """The output should be formatted as a JSON instance that conforms to the JSON schema below. As an example, for the schema {{"properties": {{"foo": {{"title": "Foo", "description": "a list of strings", "type": "array", "items": {{"type": "string"}}}}}}, "required": ["foo"]}} diff --git a/libs/core/langchain_core/output_parsers/string.py b/libs/core/langchain_core/output_parsers/string.py index 33bfffb8f7a..4f952c68892 100644 --- a/libs/core/langchain_core/output_parsers/string.py +++ b/libs/core/langchain_core/output_parsers/string.py @@ -31,6 +31,3 @@ class StrOutputParser(BaseTransformOutputParser[str]): def parse(self, text: str) -> str: """Returns the input text with no changes.""" return text - - -StrOutputParser.model_rebuild() diff --git a/libs/core/langchain_core/prompts/pipeline.py b/libs/core/langchain_core/prompts/pipeline.py index 8ec5cee9459..3c771b2dab0 100644 --- a/libs/core/langchain_core/prompts/pipeline.py +++ b/libs/core/langchain_core/prompts/pipeline.py @@ -132,6 +132,3 @@ class PipelinePromptTemplate(BasePromptTemplate): @property def _prompt_type(self) -> str: raise ValueError - - -PipelinePromptTemplate.model_rebuild() diff --git a/libs/core/langchain_core/runnables/base.py b/libs/core/langchain_core/runnables/base.py index 3b13ebcab4c..e51f1d709fd 100644 --- a/libs/core/langchain_core/runnables/base.py +++ b/libs/core/langchain_core/runnables/base.py @@ -5650,9 +5650,6 @@ class RunnableBindingBase(RunnableSerializable[Input, Output]): yield item -RunnableBindingBase.model_rebuild() - - class RunnableBinding(RunnableBindingBase[Input, Output]): """Wrap a Runnable with additional functionality. diff --git a/libs/core/langchain_core/runnables/configurable.py b/libs/core/langchain_core/runnables/configurable.py index 22a1f846d2b..0fdb2869411 100644 --- a/libs/core/langchain_core/runnables/configurable.py +++ b/libs/core/langchain_core/runnables/configurable.py @@ -8,7 +8,6 @@ from abc import abstractmethod from collections.abc import ( AsyncIterator, Iterator, - Mapping, # noqa: F401 Needed by pydantic Sequence, ) from functools import wraps @@ -464,9 +463,6 @@ class RunnableConfigurableFields(DynamicRunnable[Input, Output]): return (self.default, config) -RunnableConfigurableFields.model_rebuild() - - # Before Python 3.11 native StrEnum is not available class StrEnum(str, enum.Enum): """String enum.""" diff --git a/libs/core/langchain_core/runnables/utils.py b/libs/core/langchain_core/runnables/utils.py index c0fdeb96ae0..b9fd1b20885 100644 --- a/libs/core/langchain_core/runnables/utils.py +++ b/libs/core/langchain_core/runnables/utils.py @@ -6,6 +6,7 @@ import ast import asyncio import inspect import textwrap +from collections.abc import Mapping, Sequence from contextvars import Context from functools import lru_cache from inspect import signature @@ -33,8 +34,6 @@ if TYPE_CHECKING: Awaitable, Coroutine, Iterable, - Mapping, - Sequence, ) from langchain_core.runnables.schema import StreamEvent diff --git a/libs/core/langchain_core/tools/simple.py b/libs/core/langchain_core/tools/simple.py index e0062368f3b..05dc1917644 100644 --- a/libs/core/langchain_core/tools/simple.py +++ b/libs/core/langchain_core/tools/simple.py @@ -176,6 +176,3 @@ class Tool(BaseTool): args_schema=args_schema, **kwargs, ) - - -Tool.model_rebuild() diff --git a/libs/core/tests/unit_tests/language_models/chat_models/test_rate_limiting.py b/libs/core/tests/unit_tests/language_models/chat_models/test_rate_limiting.py index e9a72d61c6f..c4d6a50f6be 100644 --- a/libs/core/tests/unit_tests/language_models/chat_models/test_rate_limiting.py +++ b/libs/core/tests/unit_tests/language_models/chat_models/test_rate_limiting.py @@ -227,9 +227,6 @@ class SerializableModel(GenericFakeChatModel): return True -SerializableModel.model_rebuild() - - def test_serialization_with_rate_limiter() -> None: """Test model serialization with rate limiter.""" from langchain_core.load import dumps diff --git a/libs/core/tests/unit_tests/output_parsers/test_base_parsers.py b/libs/core/tests/unit_tests/output_parsers/test_base_parsers.py index 94afd7afb49..fa5e9c9c9c0 100644 --- a/libs/core/tests/unit_tests/output_parsers/test_base_parsers.py +++ b/libs/core/tests/unit_tests/output_parsers/test_base_parsers.py @@ -45,8 +45,6 @@ def test_base_generation_parser() -> None: assert isinstance(content, str) return content.swapcase() - StrInvertCase.model_rebuild() - model = GenericFakeChatModel(messages=iter([AIMessage(content="hEllo")])) chain = model | StrInvertCase() assert chain.invoke("") == "HeLLO" diff --git a/libs/core/tests/unit_tests/prompts/test_structured.py b/libs/core/tests/unit_tests/prompts/test_structured.py index 6e69936223a..8f04f2029ba 100644 --- a/libs/core/tests/unit_tests/prompts/test_structured.py +++ b/libs/core/tests/unit_tests/prompts/test_structured.py @@ -35,9 +35,6 @@ class FakeStructuredChatModel(FakeListChatModel): return "fake-messages-list-chat-model" -FakeStructuredChatModel.model_rebuild() - - def test_structured_prompt_pydantic() -> None: class OutputSchema(BaseModel): name: str diff --git a/libs/core/tests/unit_tests/runnables/test_runnable_events_v1.py b/libs/core/tests/unit_tests/runnables/test_runnable_events_v1.py index 824cbbe9056..a046070e7eb 100644 --- a/libs/core/tests/unit_tests/runnables/test_runnable_events_v1.py +++ b/libs/core/tests/unit_tests/runnables/test_runnable_events_v1.py @@ -1188,9 +1188,6 @@ class HardCodedRetriever(BaseRetriever): return self.documents -HardCodedRetriever.model_rebuild() - - async def test_event_stream_with_retriever() -> None: """Test the event stream with a retriever.""" retriever = HardCodedRetriever( diff --git a/libs/core/tests/unit_tests/test_pydantic_imports.py b/libs/core/tests/unit_tests/test_pydantic_imports.py new file mode 100644 index 00000000000..523647a7b7a --- /dev/null +++ b/libs/core/tests/unit_tests/test_pydantic_imports.py @@ -0,0 +1,20 @@ +import importlib +from pathlib import Path + +from pydantic import BaseModel + + +def test_all_models_built() -> None: + for path in Path("../core/langchain_core/").glob("*"): + module_name = path.stem + if not module_name.startswith(".") and path.suffix != ".typed": + module = importlib.import_module("langchain_core." + module_name) + all_ = getattr(module, "__all__", []) + for attr_name in all_: + attr = getattr(module, attr_name) + try: + if issubclass(attr, BaseModel): + assert attr.__pydantic_complete__ is True + except TypeError: + # This is expected for non-class attributes + pass diff --git a/libs/core/tests/unit_tests/test_tools.py b/libs/core/tests/unit_tests/test_tools.py index ca6de8f671d..3ffc15ad0fd 100644 --- a/libs/core/tests/unit_tests/test_tools.py +++ b/libs/core/tests/unit_tests/test_tools.py @@ -1091,9 +1091,6 @@ class FooBase(BaseTool): return assert_bar(bar, bar_config) -FooBase.model_rebuild() - - class AFooBase(FooBase): async def _arun(self, bar: Any, bar_config: RunnableConfig, **kwargs: Any) -> Any: return assert_bar(bar, bar_config) diff --git a/uv.lock b/uv.lock index c5cf84623ba..77b81f4a29b 100644 --- a/uv.lock +++ b/uv.lock @@ -2338,7 +2338,7 @@ dependencies = [ requires-dist = [ { name = "chromadb", specifier = ">=0.4.0,!=0.5.4,!=0.5.5,!=0.5.7,!=0.5.9,!=0.5.10,!=0.5.11,!=0.5.12,<0.7.0" }, { name = "langchain-core", editable = "libs/core" }, - { name = "numpy", marker = "python_full_version < '3.13'", specifier = ">=1.22.4" }, + { name = "numpy", marker = "python_full_version < '3.13'", specifier = ">=1.26.0" }, { name = "numpy", marker = "python_full_version >= '3.13'", specifier = ">=2.1.0" }, ] @@ -2451,7 +2451,7 @@ typing = [ { name = "langchain", editable = "libs/langchain" }, { name = "langchain-core", editable = "libs/core" }, { name = "langchain-text-splitters", editable = "libs/text-splitters" }, - { name = "mypy", specifier = ">=1.12,<2.0" }, + { name = "mypy", specifier = ">=1.15,<2.0" }, { name = "mypy-protobuf", specifier = ">=3.0.0,<4.0.0" }, { name = "types-chardet", specifier = ">=5.0.4.6,<6.0.0.0" }, { name = "types-pytz", specifier = ">=2023.3.0.0,<2024.0.0.0" }, @@ -2503,6 +2503,8 @@ test = [ { name = "numpy", marker = "python_full_version >= '3.13'", specifier = ">=2.1.0" }, { name = "pytest", specifier = ">=8,<9" }, { name = "pytest-asyncio", specifier = ">=0.21.1,<1.0.0" }, + { name = "pytest-benchmark" }, + { name = "pytest-codspeed" }, { name = "pytest-mock", specifier = ">=3.10.0,<4.0.0" }, { name = "pytest-socket", specifier = ">=0.7.0,<1.0.0" }, { name = "pytest-watcher", specifier = ">=0.3.4,<1.0.0" }, @@ -2513,8 +2515,7 @@ test = [ test-integration = [] typing = [ { name = "langchain-text-splitters", directory = "libs/text-splitters" }, - { name = "mypy", specifier = ">=1.10,<1.11" }, - { name = "types-jinja2", specifier = ">=2.11.9,<3.0.0" }, + { name = "mypy", specifier = ">=1.15,<1.16" }, { name = "types-pyyaml", specifier = ">=6.0.12.2,<7.0.0.0" }, { name = "types-requests", specifier = ">=2.28.11.5,<3.0.0.0" }, ] From fa362189a152f76b839a1b950ab840d8b1b33069 Mon Sep 17 00:00:00 2001 From: ccurme Date: Wed, 16 Apr 2025 15:21:14 -0400 Subject: [PATCH 5/7] docs: document OpenAI reasoning summaries (#30882) --- docs/docs/integrations/chat/openai.ipynb | 73 ++++++++++++++++++- .../langchain_openai/chat_models/base.py | 34 +++++++++ 2 files changed, 106 insertions(+), 1 deletion(-) diff --git a/docs/docs/integrations/chat/openai.ipynb b/docs/docs/integrations/chat/openai.ipynb index cb60106cf26..76a6f9b191f 100644 --- a/docs/docs/integrations/chat/openai.ipynb +++ b/docs/docs/integrations/chat/openai.ipynb @@ -408,7 +408,7 @@ "\n", ":::\n", "\n", - "OpenAI supports a [Responses](https://platform.openai.com/docs/guides/responses-vs-chat-completions) API that is oriented toward building [agentic](/docs/concepts/agents/) applications. It includes a suite of [built-in tools](https://platform.openai.com/docs/guides/tools?api-mode=responses), including web and file search. It also supports management of [conversation state](https://platform.openai.com/docs/guides/conversation-state?api-mode=responses), allowing you to continue a conversational thread without explicitly passing in previous messages.\n", + "OpenAI supports a [Responses](https://platform.openai.com/docs/guides/responses-vs-chat-completions) API that is oriented toward building [agentic](/docs/concepts/agents/) applications. It includes a suite of [built-in tools](https://platform.openai.com/docs/guides/tools?api-mode=responses), including web and file search. It also supports management of [conversation state](https://platform.openai.com/docs/guides/conversation-state?api-mode=responses), allowing you to continue a conversational thread without explicitly passing in previous messages, as well as the output from [reasoning processes](https://platform.openai.com/docs/guides/reasoning?api-mode=responses).\n", "\n", "`ChatOpenAI` will route to the Responses API if one of these features is used. You can also specify `use_responses_api=True` when instantiating `ChatOpenAI`.\n", "\n", @@ -1056,6 +1056,77 @@ "print(second_response.text())" ] }, + { + "cell_type": "markdown", + "id": "67bf5bd2-0935-40a0-b1cd-c6662b681d4b", + "metadata": {}, + "source": [ + "### Reasoning output\n", + "\n", + "Some OpenAI models will generate separate text content illustrating their reasoning process. See OpenAI's [reasoning documentation](https://platform.openai.com/docs/guides/reasoning?api-mode=responses) for details.\n", + "\n", + "OpenAI can return a summary of the model's reasoning (although it doesn't expose the raw reasoning tokens). To configure `ChatOpenAI` to return this summary, specify the `reasoning` parameter:" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "8d322f3a-0732-45ab-ac95-dfd4596e0d85", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Output: 3^3 = 27.\n", + "\n", + "Reasoning: {'id': 'rs_67fffc44b1c08191b6ca9bead6d832590433145b1786f809', 'summary': [{'text': '**Calculating 3 cubed**\\n\\nThe user wants to know the value of 3 to the power of 3. That\\'s straightforward—it equals 27. I can respond simply: \"3^3 equals 27.\" To provide a little clarity, I might add that it comes from multiplying 3 by itself twice: 3 * 3 * 3 = 27. So ultimately, my answer to the user is: 27. I think keeping it plain and simple is the best approach here!', 'type': 'summary_text'}], 'type': 'reasoning'}\n" + ] + } + ], + "source": [ + "from langchain_openai import ChatOpenAI\n", + "\n", + "reasoning = {\n", + " \"effort\": \"medium\", # 'low', 'medium', or 'high'\n", + " \"summary\": \"auto\", # 'detailed', 'auto', or None\n", + "}\n", + "\n", + "llm = ChatOpenAI(\n", + " model=\"o4-mini\",\n", + " use_responses_api=True,\n", + " model_kwargs={\"reasoning\": reasoning},\n", + ")\n", + "response = llm.invoke(\"What is 3^3?\")\n", + "\n", + "print(f\"Output: {response.text()}\\n\")\n", + "print(f\"Reasoning: {response.additional_kwargs['reasoning']}\")" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "id": "d7dcc082-b7c8-41b7-a5e2-441b9679e41b", + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "{'id': 'rs_67fffc44b1c08191b6ca9bead6d832590433145b1786f809',\n", + " 'summary': [{'text': '**Calculating 3 cubed**\\n\\nThe user wants to know the value of 3 to the power of 3. That\\'s straightforward—it equals 27. I can respond simply: \"3^3 equals 27.\" To provide a little clarity, I might add that it comes from multiplying 3 by itself twice: 3 * 3 * 3 = 27. So ultimately, my answer to the user is: 27. I think keeping it plain and simple is the best approach here!',\n", + " 'type': 'summary_text'}],\n", + " 'type': 'reasoning'}" + ] + }, + "execution_count": 4, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "response.additional_kwargs[\"reasoning\"]" + ] + }, { "cell_type": "markdown", "id": "57e27714", diff --git a/libs/partners/openai/langchain_openai/chat_models/base.py b/libs/partners/openai/langchain_openai/chat_models/base.py index 6670a3b7668..29fb8e9cde8 100644 --- a/libs/partners/openai/langchain_openai/chat_models/base.py +++ b/libs/partners/openai/langchain_openai/chat_models/base.py @@ -2132,6 +2132,40 @@ class ChatOpenAI(BaseChatOpenAI): # type: ignore[override] "Your name is Bob. How can I help you today, Bob?" + .. dropdown:: Reasoning output + + OpenAI's Responses API supports `reasoning models `_ + that expose a summary of internal reasoning processes. + + .. code-block:: python + + from langchain_openai import ChatOpenAI + + reasoning = { + "effort": "medium", # 'low', 'medium', or 'high' + "summary": "auto", # 'detailed', 'auto', or None + } + + llm = ChatOpenAI( + model="o4-mini", use_responses_api=True, model_kwargs={"reasoning": reasoning} + ) + response = llm.invoke("What is 3^3?") + + print(f"Output: {response.text()}") + print(f"Reasoning: {response.additional_kwargs['reasoning']}") + + .. code-block:: none + + Output: 3^3 = 27. + + Reasoning: { + 'id': 'rs_67fffc44b1c08191b6ca9bead6d832590433145b1786f809', + 'summary': [ + {'text': 'The user wants to know...', 'type': 'summary_text'} + ], + 'type': 'reasoning' + } + .. dropdown:: Structured output .. code-block:: python From 9e82f1df4e4bbedcc7bd61fdcd001432de61aedc Mon Sep 17 00:00:00 2001 From: ccurme Date: Wed, 16 Apr 2025 16:08:43 -0400 Subject: [PATCH 6/7] docs: minor clean up in ChatOpenAI docs (#30884) --- docs/docs/integrations/chat/openai.ipynb | 46 ++++++++++++------------ 1 file changed, 23 insertions(+), 23 deletions(-) diff --git a/docs/docs/integrations/chat/openai.ipynb b/docs/docs/integrations/chat/openai.ipynb index 76a6f9b191f..457293aac9e 100644 --- a/docs/docs/integrations/chat/openai.ipynb +++ b/docs/docs/integrations/chat/openai.ipynb @@ -1070,18 +1070,19 @@ }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 2, "id": "8d322f3a-0732-45ab-ac95-dfd4596e0d85", "metadata": {}, "outputs": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "Output: 3^3 = 27.\n", - "\n", - "Reasoning: {'id': 'rs_67fffc44b1c08191b6ca9bead6d832590433145b1786f809', 'summary': [{'text': '**Calculating 3 cubed**\\n\\nThe user wants to know the value of 3 to the power of 3. That\\'s straightforward—it equals 27. I can respond simply: \"3^3 equals 27.\" To provide a little clarity, I might add that it comes from multiplying 3 by itself twice: 3 * 3 * 3 = 27. So ultimately, my answer to the user is: 27. I think keeping it plain and simple is the best approach here!', 'type': 'summary_text'}], 'type': 'reasoning'}\n" - ] + "data": { + "text/plain": [ + "'3^3 = 3 × 3 × 3 = 27.'" + ] + }, + "execution_count": 2, + "metadata": {}, + "output_type": "execute_result" } ], "source": [ @@ -1099,32 +1100,31 @@ ")\n", "response = llm.invoke(\"What is 3^3?\")\n", "\n", - "print(f\"Output: {response.text()}\\n\")\n", - "print(f\"Reasoning: {response.additional_kwargs['reasoning']}\")" + "# Output\n", + "response.text()" ] }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 3, "id": "d7dcc082-b7c8-41b7-a5e2-441b9679e41b", "metadata": {}, "outputs": [ { - "data": { - "text/plain": [ - "{'id': 'rs_67fffc44b1c08191b6ca9bead6d832590433145b1786f809',\n", - " 'summary': [{'text': '**Calculating 3 cubed**\\n\\nThe user wants to know the value of 3 to the power of 3. That\\'s straightforward—it equals 27. I can respond simply: \"3^3 equals 27.\" To provide a little clarity, I might add that it comes from multiplying 3 by itself twice: 3 * 3 * 3 = 27. So ultimately, my answer to the user is: 27. I think keeping it plain and simple is the best approach here!',\n", - " 'type': 'summary_text'}],\n", - " 'type': 'reasoning'}" - ] - }, - "execution_count": 4, - "metadata": {}, - "output_type": "execute_result" + "name": "stdout", + "output_type": "stream", + "text": [ + "**Calculating power of three**\n", + "\n", + "The user is asking for the result of 3 to the power of 3, which I know is 27. It's a straightforward question, so I’ll keep my answer concise: 27. I could explain that this is the same as multiplying 3 by itself twice: 3 × 3 × 3 equals 27. However, since the user likely just needs the answer, I’ll simply respond with 27.\n" + ] } ], "source": [ - "response.additional_kwargs[\"reasoning\"]" + "# Reasoning\n", + "reasoning = response.additional_kwargs[\"reasoning\"]\n", + "for block in reasoning[\"summary\"]:\n", + " print(block[\"text\"])" ] }, { From b36c2bf833a33263bfcf188b302db8f6546f8991 Mon Sep 17 00:00:00 2001 From: ccurme Date: Wed, 16 Apr 2025 16:55:14 -0400 Subject: [PATCH 7/7] docs: update Bedrock chat model page (#30883) - document prompt caching - feature ChatBedrockConverse throughout --- docs/docs/integrations/chat/bedrock.ipynb | 773 +++++++++++----------- docs/docs/integrations/providers/aws.mdx | 13 +- 2 files changed, 409 insertions(+), 377 deletions(-) diff --git a/docs/docs/integrations/chat/bedrock.ipynb b/docs/docs/integrations/chat/bedrock.ipynb index bcc916af098..d69a7b22cc1 100644 --- a/docs/docs/integrations/chat/bedrock.ipynb +++ b/docs/docs/integrations/chat/bedrock.ipynb @@ -1,378 +1,401 @@ { - "cells": [ - { - "cell_type": "raw", - "id": "afaf8039", - "metadata": {}, - "source": [ - "---\n", - "sidebar_label: AWS Bedrock\n", - "---" - ] - }, - { - "cell_type": "markdown", - "id": "e49f1e0d", - "metadata": {}, - "source": [ - "# ChatBedrock\n", - "\n", - "This doc will help you get started with AWS Bedrock [chat models](/docs/concepts/chat_models). Amazon Bedrock is a fully managed service that offers a choice of high-performing foundation models (FMs) from leading AI companies like AI21 Labs, Anthropic, Cohere, Meta, Stability AI, and Amazon via a single API, along with a broad set of capabilities you need to build generative AI applications with security, privacy, and responsible AI. Using Amazon Bedrock, you can easily experiment with and evaluate top FMs for your use case, privately customize them with your data using techniques such as fine-tuning and Retrieval Augmented Generation (RAG), and build agents that execute tasks using your enterprise systems and data sources. Since Amazon Bedrock is serverless, you don't have to manage any infrastructure, and you can securely integrate and deploy generative AI capabilities into your applications using the AWS services you are already familiar with.\n", - "\n", - "For more information on which models are accessible via Bedrock, head to the [AWS docs](https://docs.aws.amazon.com/bedrock/latest/userguide/models-features.html).\n", - "\n", - "For detailed documentation of all ChatBedrock features and configurations head to the [API reference](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock.ChatBedrock.html).\n", - "\n", - "## Overview\n", - "### Integration details\n", - "\n", - "| Class | Package | Local | Serializable | [JS support](https://js.langchain.com/docs/integrations/chat/bedrock) | Package downloads | Package latest |\n", - "| :--- | :--- | :---: | :---: | :---: | :---: | :---: |\n", - "| [ChatBedrock](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock.ChatBedrock.html) | [langchain-aws](https://python.langchain.com/api_reference/aws/index.html) | ❌ | beta | ✅ | ![PyPI - Downloads](https://img.shields.io/pypi/dm/langchain-aws?style=flat-square&label=%20) | ![PyPI - Version](https://img.shields.io/pypi/v/langchain-aws?style=flat-square&label=%20) |\n", - "\n", - "### Model features\n", - "| [Tool calling](/docs/how_to/tool_calling) | [Structured output](/docs/how_to/structured_output/) | JSON mode | [Image input](/docs/how_to/multimodal_inputs/) | Audio input | Video input | [Token-level streaming](/docs/how_to/chat_streaming/) | Native async | [Token usage](/docs/how_to/chat_token_usage_tracking/) | [Logprobs](/docs/how_to/logprobs/) |\n", - "| :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |\n", - "| ✅ | ✅ | ❌ | ✅ | ❌ | ❌ | ✅ | ❌ | ✅ | ❌ |\n", - "\n", - "## Setup\n", - "\n", - "To access Bedrock models you'll need to create an AWS account, set up the Bedrock API service, get an access key ID and secret key, and install the `langchain-aws` integration package.\n", - "\n", - "### Credentials\n", - "\n", - "Head to the [AWS docs](https://docs.aws.amazon.com/bedrock/latest/userguide/setting-up.html) to sign up to AWS and setup your credentials. You'll also need to turn on model access for your account, which you can do by following [these instructions](https://docs.aws.amazon.com/bedrock/latest/userguide/model-access.html)." - ] - }, - { - "cell_type": "markdown", - "id": "72ee0c4b-9764-423a-9dbf-95129e185210", - "metadata": {}, - "source": "To enable automated tracing of your model calls, set your [LangSmith](https://docs.smith.langchain.com/) API key:" - }, - { - "cell_type": "code", - "execution_count": null, - "id": "a15d341e-3e26-4ca3-830b-5aab30ed66de", - "metadata": {}, - "outputs": [], - "source": [ - "# os.environ[\"LANGSMITH_API_KEY\"] = getpass.getpass(\"Enter your LangSmith API key: \")\n", - "# os.environ[\"LANGSMITH_TRACING\"] = \"true\"" - ] - }, - { - "cell_type": "markdown", - "id": "0730d6a1-c893-4840-9817-5e5251676d5d", - "metadata": {}, - "source": [ - "### Installation\n", - "\n", - "The LangChain Bedrock integration lives in the `langchain-aws` package:" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "id": "652d6238-1f87-422a-b135-f5abbb8652fc", - "metadata": {}, - "outputs": [], - "source": [ - "%pip install -qU langchain-aws" - ] - }, - { - "cell_type": "markdown", - "id": "a38cde65-254d-4219-a441-068766c0d4b5", - "metadata": {}, - "source": [ - "## Instantiation\n", - "\n", - "Now we can instantiate our model object and generate chat completions:" - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "id": "cb09c344-1836-4e0c-acf8-11d13ac1dbae", - "metadata": {}, - "outputs": [], - "source": [ - "from langchain_aws import ChatBedrock\n", - "\n", - "llm = ChatBedrock(\n", - " model_id=\"anthropic.claude-3-sonnet-20240229-v1:0\",\n", - " model_kwargs=dict(temperature=0),\n", - " # other params...\n", - ")" - ] - }, - { - "cell_type": "markdown", - "id": "2b4f3e15", - "metadata": {}, - "source": [ - "## Invocation" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "id": "62e0dbc3", - "metadata": { - "tags": [] - }, - "outputs": [ - { - "data": { - "text/plain": [ - "AIMessage(content=\"Voici la traduction en français :\\n\\nJ'aime la programmation.\", additional_kwargs={'usage': {'prompt_tokens': 29, 'completion_tokens': 21, 'total_tokens': 50}, 'stop_reason': 'end_turn', 'model_id': 'anthropic.claude-3-sonnet-20240229-v1:0'}, response_metadata={'usage': {'prompt_tokens': 29, 'completion_tokens': 21, 'total_tokens': 50}, 'stop_reason': 'end_turn', 'model_id': 'anthropic.claude-3-sonnet-20240229-v1:0'}, id='run-fdb07dc3-ff72-430d-b22b-e7824b15c766-0', usage_metadata={'input_tokens': 29, 'output_tokens': 21, 'total_tokens': 50})" - ] - }, - "execution_count": 5, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "messages = [\n", - " (\n", - " \"system\",\n", - " \"You are a helpful assistant that translates English to French. Translate the user sentence.\",\n", - " ),\n", - " (\"human\", \"I love programming.\"),\n", - "]\n", - "ai_msg = llm.invoke(messages)\n", - "ai_msg" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "id": "d86145b3-bfef-46e8-b227-4dda5c9c2705", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Voici la traduction en français :\n", - "\n", - "J'aime la programmation.\n" - ] - } - ], - "source": [ - "print(ai_msg.content)" - ] - }, - { - "cell_type": "markdown", - "id": "18e2bfc0-7e78-4528-a73f-499ac150dca8", - "metadata": {}, - "source": [ - "## Chaining\n", - "\n", - "We can [chain](/docs/how_to/sequence/) our model with a prompt template like so:" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "id": "e197d1d7-a070-4c96-9f8a-a0e86d046e0b", - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "AIMessage(content='Ich liebe Programmieren.', additional_kwargs={'usage': {'prompt_tokens': 23, 'completion_tokens': 11, 'total_tokens': 34}, 'stop_reason': 'end_turn', 'model_id': 'anthropic.claude-3-sonnet-20240229-v1:0'}, response_metadata={'usage': {'prompt_tokens': 23, 'completion_tokens': 11, 'total_tokens': 34}, 'stop_reason': 'end_turn', 'model_id': 'anthropic.claude-3-sonnet-20240229-v1:0'}, id='run-5ad005ce-9f31-4670-baa0-9373d418698a-0', usage_metadata={'input_tokens': 23, 'output_tokens': 11, 'total_tokens': 34})" - ] - }, - "execution_count": 7, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "from langchain_core.prompts import ChatPromptTemplate\n", - "\n", - "prompt = ChatPromptTemplate.from_messages(\n", - " [\n", - " (\n", - " \"system\",\n", - " \"You are a helpful assistant that translates {input_language} to {output_language}.\",\n", - " ),\n", - " (\"human\", \"{input}\"),\n", - " ]\n", - ")\n", - "\n", - "chain = prompt | llm\n", - "chain.invoke(\n", - " {\n", - " \"input_language\": \"English\",\n", - " \"output_language\": \"German\",\n", - " \"input\": \"I love programming.\",\n", - " }\n", - ")" - ] - }, - { - "cell_type": "markdown", - "id": "d1ee55bc-ffc8-4cfa-801c-993953a08cfd", - "metadata": {}, - "source": [ - "## Bedrock Converse API\n", - "\n", - "AWS has recently released the Bedrock Converse API which provides a unified conversational interface for Bedrock models. This API does not yet support custom models. You can see a list of all [models that are supported here](https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html). To improve reliability the ChatBedrock integration will switch to using the Bedrock Converse API as soon as it has feature parity with the existing Bedrock API. Until then a separate [ChatBedrockConverse](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock_converse.ChatBedrockConverse.html) integration has been released.\n", - "\n", - "We recommend using `ChatBedrockConverse` for users who do not need to use custom models.\n", - "\n", - "You can use it like so:" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "id": "ae728e59-94d4-40cf-9d24-25ad8723fc59", - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "AIMessage(content=\"Voici la traduction en français :\\n\\nJ'aime la programmation.\", response_metadata={'ResponseMetadata': {'RequestId': '4fcbfbe9-f916-4df2-b0bd-ea1147b550aa', 'HTTPStatusCode': 200, 'HTTPHeaders': {'date': 'Wed, 21 Aug 2024 17:23:49 GMT', 'content-type': 'application/json', 'content-length': '243', 'connection': 'keep-alive', 'x-amzn-requestid': '4fcbfbe9-f916-4df2-b0bd-ea1147b550aa'}, 'RetryAttempts': 0}, 'stopReason': 'end_turn', 'metrics': {'latencyMs': 672}}, id='run-77ee9810-e32b-45dc-9ccb-6692253b1f45-0', usage_metadata={'input_tokens': 29, 'output_tokens': 21, 'total_tokens': 50})" - ] - }, - "execution_count": 3, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "from langchain_aws import ChatBedrockConverse\n", - "\n", - "llm = ChatBedrockConverse(\n", - " model=\"anthropic.claude-3-sonnet-20240229-v1:0\",\n", - " temperature=0,\n", - " max_tokens=None,\n", - " # other params...\n", - ")\n", - "\n", - "llm.invoke(messages)" - ] - }, - { - "cell_type": "markdown", - "id": "4da16f3e-e80b-48c0-8036-c1cc5f7c8c05", - "metadata": {}, - "source": [ - "### Streaming\n", - "\n", - "Note that `ChatBedrockConverse` emits content blocks while streaming:" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "id": "7794b32e-d8de-4973-bf0f-39807dc745f0", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "content=[] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': 'Vo', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': 'ici', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': ' la', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': ' tra', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': 'duction', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': ' en', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': ' français', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': ' :', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': '\\n\\nJ', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': \"'\", 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': 'a', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': 'ime', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': ' la', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': ' programm', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': 'ation', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'type': 'text', 'text': '.', 'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[{'index': 0}] id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[] response_metadata={'stopReason': 'end_turn'} id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8'\n", - "content=[] response_metadata={'metrics': {'latencyMs': 713}} id='run-2c92c5af-d771-4cc2-98d9-c11bbd30a1d8' usage_metadata={'input_tokens': 29, 'output_tokens': 21, 'total_tokens': 50}\n" - ] - } - ], - "source": [ - "for chunk in llm.stream(messages):\n", - " print(chunk)" - ] - }, - { - "cell_type": "markdown", - "id": "0ef05abb-9c04-4dc3-995e-f857779644d5", - "metadata": {}, - "source": [ - "An output parser can be used to filter to text, if desired:" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "id": "2a4e743f-ea7d-4e5a-9b12-f9992362de8b", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "|Vo|ici| la| tra|duction| en| français| :|\n", - "\n", - "J|'|a|ime| la| programm|ation|.||||" - ] - } - ], - "source": [ - "from langchain_core.output_parsers import StrOutputParser\n", - "\n", - "chain = llm | StrOutputParser()\n", - "\n", - "for chunk in chain.stream(messages):\n", - " print(chunk, end=\"|\")" - ] - }, - { - "cell_type": "markdown", - "id": "3a5bb5ca-c3ae-4a58-be67-2cd18574b9a3", - "metadata": {}, - "source": [ - "## API reference\n", - "\n", - "For detailed documentation of all ChatBedrock features and configurations head to the API reference: https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock.ChatBedrock.html\n", - "\n", - "For detailed documentation of all ChatBedrockConverse features and configurations head to the API reference: https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock_converse.ChatBedrockConverse.html" - ] - } - ], - "metadata": { - "kernelspec": { - "display_name": "Python 3 (ipykernel)", - "language": "python", - "name": "python3" - }, - "language_info": { - "codemirror_mode": { - "name": "ipython", - "version": 3 - }, - "file_extension": ".py", - "mimetype": "text/x-python", - "name": "python", - "nbconvert_exporter": "python", - "pygments_lexer": "ipython3", - "version": "3.10.4" - } + "cells": [ + { + "cell_type": "raw", + "id": "afaf8039", + "metadata": {}, + "source": [ + "---\n", + "sidebar_label: AWS Bedrock\n", + "---" + ] }, - "nbformat": 4, - "nbformat_minor": 5 + { + "cell_type": "markdown", + "id": "e49f1e0d", + "metadata": {}, + "source": [ + "# ChatBedrock\n", + "\n", + "This doc will help you get started with AWS Bedrock [chat models](/docs/concepts/chat_models). Amazon Bedrock is a fully managed service that offers a choice of high-performing foundation models (FMs) from leading AI companies like AI21 Labs, Anthropic, Cohere, Meta, Stability AI, and Amazon via a single API, along with a broad set of capabilities you need to build generative AI applications with security, privacy, and responsible AI. Using Amazon Bedrock, you can easily experiment with and evaluate top FMs for your use case, privately customize them with your data using techniques such as fine-tuning and Retrieval Augmented Generation (RAG), and build agents that execute tasks using your enterprise systems and data sources. Since Amazon Bedrock is serverless, you don't have to manage any infrastructure, and you can securely integrate and deploy generative AI capabilities into your applications using the AWS services you are already familiar with.\n", + "\n", + "AWS Bedrock maintains a [Converse API](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_Converse.html) which provides a unified conversational interface for Bedrock models. This API does not yet support custom models. You can see a list of all [models that are supported here](https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html).\n", + "\n", + ":::info\n", + "\n", + "We recommend the Converse API for users who do not need to use custom models. It can be accessed using [ChatBedrockConverse](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock_converse.ChatBedrockConverse.html).\n", + "\n", + ":::\n", + "\n", + "For detailed documentation of all Bedrock features and configurations head to the [API reference](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock_converse.ChatBedrockConverse.html).\n", + "\n", + "## Overview\n", + "### Integration details\n", + "\n", + "| Class | Package | Local | Serializable | [JS support](https://js.langchain.com/docs/integrations/chat/bedrock) | Package downloads | Package latest |\n", + "| :--- | :--- | :---: | :---: | :---: | :---: | :---: |\n", + "| [ChatBedrock](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock.ChatBedrock.html) | [langchain-aws](https://python.langchain.com/api_reference/aws/index.html) | ❌ | beta | ✅ | ![PyPI - Downloads](https://img.shields.io/pypi/dm/langchain-aws?style=flat-square&label=%20) | ![PyPI - Version](https://img.shields.io/pypi/v/langchain-aws?style=flat-square&label=%20) |\n", + "| [ChatBedrockConverse](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock_converse.ChatBedrockConverse.html) | [langchain-aws](https://python.langchain.com/api_reference/aws/index.html) | ❌ | beta | ✅ | ![PyPI - Downloads](https://img.shields.io/pypi/dm/langchain-aws?style=flat-square&label=%20) | ![PyPI - Version](https://img.shields.io/pypi/v/langchain-aws?style=flat-square&label=%20) |\n", + "\n", + "### Model features\n", + "\n", + "The below apply to both `ChatBedrock` and `ChatBedrockConverse`.\n", + "\n", + "| [Tool calling](/docs/how_to/tool_calling) | [Structured output](/docs/how_to/structured_output/) | JSON mode | [Image input](/docs/how_to/multimodal_inputs/) | Audio input | Video input | [Token-level streaming](/docs/how_to/chat_streaming/) | Native async | [Token usage](/docs/how_to/chat_token_usage_tracking/) | [Logprobs](/docs/how_to/logprobs/) |\n", + "| :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |\n", + "| ✅ | ✅ | ❌ | ✅ | ❌ | ❌ | ✅ | ❌ | ✅ | ❌ |\n", + "\n", + "## Setup\n", + "\n", + "To access Bedrock models you'll need to create an AWS account, set up the Bedrock API service, get an access key ID and secret key, and install the `langchain-aws` integration package.\n", + "\n", + "### Credentials\n", + "\n", + "Head to the [AWS docs](https://docs.aws.amazon.com/bedrock/latest/userguide/setting-up.html) to sign up to AWS and setup your credentials. You'll also need to turn on model access for your account, which you can do by following [these instructions](https://docs.aws.amazon.com/bedrock/latest/userguide/model-access.html)." + ] + }, + { + "cell_type": "markdown", + "id": "72ee0c4b-9764-423a-9dbf-95129e185210", + "metadata": {}, + "source": [ + "To enable automated tracing of your model calls, set your [LangSmith](https://docs.smith.langchain.com/) API key:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "a15d341e-3e26-4ca3-830b-5aab30ed66de", + "metadata": {}, + "outputs": [], + "source": [ + "# os.environ[\"LANGSMITH_API_KEY\"] = getpass.getpass(\"Enter your LangSmith API key: \")\n", + "# os.environ[\"LANGSMITH_TRACING\"] = \"true\"" + ] + }, + { + "cell_type": "markdown", + "id": "0730d6a1-c893-4840-9817-5e5251676d5d", + "metadata": {}, + "source": [ + "### Installation\n", + "\n", + "The LangChain Bedrock integration lives in the `langchain-aws` package:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "652d6238-1f87-422a-b135-f5abbb8652fc", + "metadata": {}, + "outputs": [], + "source": [ + "%pip install -qU langchain-aws" + ] + }, + { + "cell_type": "markdown", + "id": "a38cde65-254d-4219-a441-068766c0d4b5", + "metadata": {}, + "source": [ + "## Instantiation\n", + "\n", + "Now we can instantiate our model object and generate chat completions:" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "id": "cb09c344-1836-4e0c-acf8-11d13ac1dbae", + "metadata": {}, + "outputs": [], + "source": [ + "from langchain_aws import ChatBedrockConverse\n", + "\n", + "llm = ChatBedrockConverse(\n", + " model_id=\"anthropic.claude-3-5-sonnet-20240620-v1:0\",\n", + " # temperature=...,\n", + " # max_tokens=...,\n", + " # other params...\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "2b4f3e15", + "metadata": {}, + "source": [ + "## Invocation" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "fcd8de52-4a1b-4875-b463-d41b031e06a1", + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "AIMessage(content=\"J'adore la programmation.\", additional_kwargs={}, response_metadata={'ResponseMetadata': {'RequestId': 'b07d1630-06f2-44b1-82bf-e82538dd2215', 'HTTPStatusCode': 200, 'HTTPHeaders': {'date': 'Wed, 16 Apr 2025 19:35:34 GMT', 'content-type': 'application/json', 'content-length': '206', 'connection': 'keep-alive', 'x-amzn-requestid': 'b07d1630-06f2-44b1-82bf-e82538dd2215'}, 'RetryAttempts': 0}, 'stopReason': 'end_turn', 'metrics': {'latencyMs': [488]}, 'model_name': 'anthropic.claude-3-5-sonnet-20240620-v1:0'}, id='run-d09ed928-146a-4336-b1fd-b63c9e623494-0', usage_metadata={'input_tokens': 29, 'output_tokens': 11, 'total_tokens': 40, 'input_token_details': {'cache_creation': 0, 'cache_read': 0}})" + ] + }, + "execution_count": 2, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "messages = [\n", + " (\n", + " \"system\",\n", + " \"You are a helpful assistant that translates English to French. Translate the user sentence.\",\n", + " ),\n", + " (\"human\", \"I love programming.\"),\n", + "]\n", + "ai_msg = llm.invoke(messages)\n", + "ai_msg" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "d86145b3-bfef-46e8-b227-4dda5c9c2705", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "J'adore la programmation.\n" + ] + } + ], + "source": [ + "print(ai_msg.content)" + ] + }, + { + "cell_type": "markdown", + "id": "4da16f3e-e80b-48c0-8036-c1cc5f7c8c05", + "metadata": {}, + "source": [ + "### Streaming\n", + "\n", + "Note that `ChatBedrockConverse` emits content blocks while streaming:" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "id": "605e04fa-1a76-47ac-8c92-fe128659663e", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "content=[] additional_kwargs={} response_metadata={} id='run-d0e0836e-7146-4c3d-97c7-ad23dac6febd'\n", + "content=[{'type': 'text', 'text': 'J', 'index': 0}] additional_kwargs={} response_metadata={} id='run-d0e0836e-7146-4c3d-97c7-ad23dac6febd'\n", + "content=[{'type': 'text', 'text': \"'adore la\", 'index': 0}] additional_kwargs={} response_metadata={} id='run-d0e0836e-7146-4c3d-97c7-ad23dac6febd'\n", + "content=[{'type': 'text', 'text': ' programmation.', 'index': 0}] additional_kwargs={} response_metadata={} id='run-d0e0836e-7146-4c3d-97c7-ad23dac6febd'\n", + "content=[{'index': 0}] additional_kwargs={} response_metadata={} id='run-d0e0836e-7146-4c3d-97c7-ad23dac6febd'\n", + "content=[] additional_kwargs={} response_metadata={'stopReason': 'end_turn'} id='run-d0e0836e-7146-4c3d-97c7-ad23dac6febd'\n", + "content=[] additional_kwargs={} response_metadata={'metrics': {'latencyMs': 600}, 'model_name': 'anthropic.claude-3-5-sonnet-20240620-v1:0'} id='run-d0e0836e-7146-4c3d-97c7-ad23dac6febd' usage_metadata={'input_tokens': 29, 'output_tokens': 11, 'total_tokens': 40, 'input_token_details': {'cache_creation': 0, 'cache_read': 0}}\n" + ] + } + ], + "source": [ + "for chunk in llm.stream(messages):\n", + " print(chunk)" + ] + }, + { + "cell_type": "markdown", + "id": "0ef05abb-9c04-4dc3-995e-f857779644d5", + "metadata": {}, + "source": [ + "You can filter to text using the [.text()](https://python.langchain.com/api_reference/core/messages/langchain_core.messages.ai.AIMessage.html#langchain_core.messages.ai.AIMessage.text) method on the output:" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "id": "2a4e743f-ea7d-4e5a-9b12-f9992362de8b", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "|J|'adore la| programmation.||||" + ] + } + ], + "source": [ + "for chunk in llm.stream(messages):\n", + " print(chunk.text(), end=\"|\")" + ] + }, + { + "cell_type": "markdown", + "id": "a77519e5-897d-41a0-a9bb-55300fa79efc", + "metadata": {}, + "source": [ + "## Prompt caching\n", + "\n", + "Bedrock supports [caching](https://docs.aws.amazon.com/bedrock/latest/userguide/prompt-caching.html) of elements of your prompts, including messages and tools. This allows you to re-use large documents, instructions, [few-shot documents](/docs/concepts/few_shot_prompting/), and other data to reduce latency and costs.\n", + "\n", + ":::note\n", + "\n", + "Not all models support prompt caching. See supported models [here](https://docs.aws.amazon.com/bedrock/latest/userguide/prompt-caching.html#prompt-caching-models).\n", + "\n", + ":::\n", + "\n", + "To enable caching on an element of a prompt, mark its associated content block using the `cachePoint` key. See example below:" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "id": "d5f63d01-85e8-4797-a2be-0fea747a6049", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "First invocation:\n", + "{'cache_creation': 1528, 'cache_read': 0}\n", + "\n", + "Second:\n", + "{'cache_creation': 0, 'cache_read': 1528}\n" + ] + } + ], + "source": [ + "import requests\n", + "from langchain_aws import ChatBedrockConverse\n", + "\n", + "llm = ChatBedrockConverse(model=\"us.anthropic.claude-3-7-sonnet-20250219-v1:0\")\n", + "\n", + "# Pull LangChain readme\n", + "get_response = requests.get(\n", + " \"https://raw.githubusercontent.com/langchain-ai/langchain/master/README.md\"\n", + ")\n", + "readme = get_response.text\n", + "\n", + "messages = [\n", + " {\n", + " \"role\": \"user\",\n", + " \"content\": [\n", + " {\n", + " \"type\": \"text\",\n", + " \"text\": \"What's LangChain, according to its README?\",\n", + " },\n", + " {\n", + " \"type\": \"text\",\n", + " \"text\": f\"{readme}\",\n", + " },\n", + " {\n", + " \"cachePoint\": {\"type\": \"default\"},\n", + " },\n", + " ],\n", + " },\n", + "]\n", + "\n", + "response_1 = llm.invoke(messages)\n", + "response_2 = llm.invoke(messages)\n", + "\n", + "usage_1 = response_1.usage_metadata[\"input_token_details\"]\n", + "usage_2 = response_2.usage_metadata[\"input_token_details\"]\n", + "\n", + "print(f\"First invocation:\\n{usage_1}\")\n", + "print(f\"\\nSecond:\\n{usage_2}\")" + ] + }, + { + "cell_type": "markdown", + "id": "1b550667-af5b-4557-b84f-c8f865dad6cb", + "metadata": {}, + "source": [ + "## Chaining\n", + "\n", + "We can [chain](/docs/how_to/sequence/) our model with a prompt template like so:" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "id": "6033f3fa-0e96-46e3-abb3-1530928fea88", + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "AIMessage(content=\"Here's the German translation:\\n\\nIch liebe das Programmieren.\", additional_kwargs={}, response_metadata={'ResponseMetadata': {'RequestId': '1de3d7c0-8062-4f7e-bb8a-8f725b97a8b0', 'HTTPStatusCode': 200, 'HTTPHeaders': {'date': 'Wed, 16 Apr 2025 19:32:51 GMT', 'content-type': 'application/json', 'content-length': '243', 'connection': 'keep-alive', 'x-amzn-requestid': '1de3d7c0-8062-4f7e-bb8a-8f725b97a8b0'}, 'RetryAttempts': 0}, 'stopReason': 'end_turn', 'metrics': {'latencyMs': [719]}, 'model_name': 'anthropic.claude-3-5-sonnet-20240620-v1:0'}, id='run-7021fcd7-704e-496b-a92e-210139614402-0', usage_metadata={'input_tokens': 23, 'output_tokens': 19, 'total_tokens': 42, 'input_token_details': {'cache_creation': 0, 'cache_read': 0}})" + ] + }, + "execution_count": 7, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "from langchain_core.prompts import ChatPromptTemplate\n", + "\n", + "prompt = ChatPromptTemplate.from_messages(\n", + " [\n", + " (\n", + " \"system\",\n", + " \"You are a helpful assistant that translates {input_language} to {output_language}.\",\n", + " ),\n", + " (\"human\", \"{input}\"),\n", + " ]\n", + ")\n", + "\n", + "chain = prompt | llm\n", + "chain.invoke(\n", + " {\n", + " \"input_language\": \"English\",\n", + " \"output_language\": \"German\",\n", + " \"input\": \"I love programming.\",\n", + " }\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "3a5bb5ca-c3ae-4a58-be67-2cd18574b9a3", + "metadata": {}, + "source": [ + "## API reference\n", + "\n", + "For detailed documentation of all ChatBedrock features and configurations head to the API reference: https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock.ChatBedrock.html\n", + "\n", + "For detailed documentation of all ChatBedrockConverse features and configurations head to the API reference: https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock_converse.ChatBedrockConverse.html" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3 (ipykernel)", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.10.4" + } + }, + "nbformat": 4, + "nbformat_minor": 5 } diff --git a/docs/docs/integrations/providers/aws.mdx b/docs/docs/integrations/providers/aws.mdx index e30cbfd81f8..c85b6eb8374 100755 --- a/docs/docs/integrations/providers/aws.mdx +++ b/docs/docs/integrations/providers/aws.mdx @@ -35,9 +35,18 @@ from langchain_aws import ChatBedrock ``` ### Bedrock Converse -AWS has recently released the Bedrock Converse API which provides a unified conversational interface for Bedrock models. This API does not yet support custom models. You can see a list of all [models that are supported here](https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html). To improve reliability the ChatBedrock integration will switch to using the Bedrock Converse API as soon as it has feature parity with the existing Bedrock API. Until then a separate [ChatBedrockConverse](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock_converse.ChatBedrockConverse.html) integration has been released. +AWS Bedrock maintains a [Converse API](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_Converse.html) +that provides a unified conversational interface for Bedrock models. This API does not +yet support custom models. You can see a list of all +[models that are supported here](https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html). -We recommend using `ChatBedrockConverse` for users who do not need to use custom models. See the [docs](/docs/integrations/chat/bedrock/#bedrock-converse-api) and [API reference](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock_converse.ChatBedrockConverse.html) for more detail. +:::info + +We recommend the Converse API for users who do not need to use custom models. It can be accessed using [ChatBedrockConverse](https://python.langchain.com/api_reference/aws/chat_models/langchain_aws.chat_models.bedrock_converse.ChatBedrockConverse.html). + +::: + +See a [usage example](/docs/integrations/chat/bedrock). ```python from langchain_aws import ChatBedrockConverse