Fix ChatAnthropic stop_sequences error (#2919) (#2920)

Note to self: Always run integration tests, even on "that last minute
change you thought would be safe" :)

---------

Co-authored-by: Mike Lambert <mike.lambert@anthropic.com>

docs/modules/models/chat/integrations/anthropic.ipynb

@@ -0,0 +1,171 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "bf733a38-db84-4363-89e2-de6735c37230",
+   "metadata": {},
+   "source": [
+    "# Anthropic\n",
+    "\n",
+    "This notebook covers how to get started with Anthropic chat models."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "d4a7c55d-b235-4ca4-a579-c90cc9570da9",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "from langchain.chat_models import ChatAnthropic\n",
+    "from langchain.prompts.chat import (\n",
+    "    ChatPromptTemplate,\n",
+    "    SystemMessagePromptTemplate,\n",
+    "    AIMessagePromptTemplate,\n",
+    "    HumanMessagePromptTemplate,\n",
+    ")\n",
+    "from langchain.schema import (\n",
+    "    AIMessage,\n",
+    "    HumanMessage,\n",
+    "    SystemMessage\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "70cf04e8-423a-4ff6-8b09-f11fb711c817",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "chat = ChatAnthropic()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "8199ef8f-eb8b-4253-9ea0-6c24a013ca4c",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "AIMessage(content=\" J'adore programmer.\", additional_kwargs={})"
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "messages = [\n",
+    "    HumanMessage(content=\"Translate this sentence from English to French. I love programming.\")\n",
+    "]\n",
+    "chat(messages)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "c361ab1e-8c0c-4206-9e3c-9d1424a12b9c",
+   "metadata": {},
+   "source": [
+    "## `ChatAnthropic` also supports async and streaming functionality:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "93a21c5c-6ef9-4688-be60-b2e1f94842fb",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "from langchain.callbacks.base import CallbackManager\n",
+    "from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "c5fac0e9-05a4-4fc1-a3b3-e5bbb24b971b",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "LLMResult(generations=[[ChatGeneration(text=\" J'aime programmer.\", generation_info=None, message=AIMessage(content=\" J'aime programmer.\", additional_kwargs={}))]], llm_output={})"
+      ]
+     },
+     "execution_count": 6,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "await chat.agenerate([messages])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "025be980-e50d-4a68-93dc-c9c7b500ce34",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " J'aime la programmation."
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "AIMessage(content=\" J'aime la programmation.\", additional_kwargs={})"
+      ]
+     },
+     "execution_count": 8,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "chat = ChatAnthropic(streaming=True, verbose=True, callback_manager=CallbackManager([StreamingStdOutCallbackHandler()]))\n",
+    "chat(messages)"
+   ]
+  }
+ ],
+ "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": "3.10.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

langchain/chat_models/__init__.py

@@ -1,5 +1,6 @@
+from langchain.chat_models.anthropic import ChatAnthropic
 from langchain.chat_models.azure_openai import AzureChatOpenAI
 from langchain.chat_models.openai import ChatOpenAI
 from langchain.chat_models.promptlayer_openai import PromptLayerChatOpenAI
 
-__all__ = ["ChatOpenAI", "AzureChatOpenAI", "PromptLayerChatOpenAI"]
+__all__ = ["ChatOpenAI", "AzureChatOpenAI", "PromptLayerChatOpenAI", "ChatAnthropic"]

langchain/chat_models/anthropic.py

@@ -1,4 +1,4 @@
-from typing import List, Optional
+from typing import Any, Dict, List, Optional
 
 from pydantic import Extra
 
@@ -26,17 +26,7 @@ class ChatAnthropic(BaseChatModel, _AnthropicCommon):
         .. code-block:: python
             import anthropic
             from langchain.llms import Anthropic
-            model = Anthropic(model="<model_name>", anthropic_api_key="my-api-key")
-
-            # Simplest invocation, automatically wrapped with HUMAN_PROMPT
-            # and AI_PROMPT.
-            response = model("What are the biggest risks facing humanity?")
-
-            # Or if you want to use the chat mode, build a few-shot-prompt, or
-            # put words in the Assistant's mouth, use HUMAN_PROMPT and AI_PROMPT:
-            raw_prompt = "What are the biggest risks facing humanity?"
-            prompt = f"{anthropic.HUMAN_PROMPT} {prompt}{anthropic.AI_PROMPT}"
-            response = model(prompt)
+            model = ChatAnthropic(model="<model_name>", anthropic_api_key="my-api-key")
     """
 
     class Config:
@@ -98,7 +88,9 @@ class ChatAnthropic(BaseChatModel, _AnthropicCommon):
         self, messages: List[BaseMessage], stop: Optional[List[str]] = None
     ) -> ChatResult:
         prompt = self._convert_messages_to_prompt(messages)
-        params = {"prompt": prompt, "stop_sequences": stop, **self._default_params}
+        params: Dict[str, Any] = {"prompt": prompt, **self._default_params}
+        if stop:
+            params["stop_sequences"] = stop
 
         if self.streaming:
             completion = ""
@@ -120,7 +112,9 @@ class ChatAnthropic(BaseChatModel, _AnthropicCommon):
         self, messages: List[BaseMessage], stop: Optional[List[str]] = None
     ) -> ChatResult:
         prompt = self._convert_messages_to_prompt(messages)
-        params = {"prompt": prompt, "stop_sequences": stop, **self._default_params}
+        params: Dict[str, Any] = {"prompt": prompt, **self._default_params}
+        if stop:
+            params["stop_sequences"] = stop
 
         if self.streaming:
             completion = ""

langchain/llms/anthropic.py

@@ -10,7 +10,7 @@ from langchain.utils import get_from_dict_or_env
 
 class _AnthropicCommon(BaseModel):
     client: Any = None  #: :meta private:
-    model: str = "claude-latest"
+    model: str = "claude-v1"
     """Model name to use."""
 
     max_tokens_to_sample: int = 256

poetry.lock

@@ -9035,4 +9035,4 @@ qdrant = ["qdrant-client"]
 [metadata]
 lock-version = "2.0"
 python-versions = ">=3.8.1,<4.0"
-content-hash = "373f68ef16e7f3d5d9cde8b81c5f261096cc537ddca4f6a36711d7215b63f226"
+content-hash = "7e343fa8e31d8fcf1023cbda592f64c05e80015c4e0e23c1d387d2e9671ce995"

pyproject.toml

@@ -36,7 +36,7 @@ pinecone-text = {version = "^0.4.2", optional = true}
 weaviate-client = {version = "^3", optional = true}
 google-api-python-client = {version = "2.70.0", optional = true}
 wolframalpha = {version = "5.0.0", optional = true}
-anthropic = {version = "^0.2.4", optional = true}
+anthropic = {version = "^0.2.6", optional = true}
 qdrant-client = {version = "^1.1.2", optional = true, python = ">=3.8.1,<3.12"}
 dataclasses-json = "^0.5.7"
 tensorflow-text = {version = "^2.11.0", optional = true, python = "^3.10, <3.12"}

tests/integration_tests/chat_models/test_anthropic.py

@@ -17,7 +17,7 @@ from tests.unit_tests.callbacks.fake_callback_handler import FakeCallbackHandler
 
 def test_anthropic_call() -> None:
     """Test valid call to anthropic."""
-    chat = ChatAnthropic(model="bare-nano-0")
+    chat = ChatAnthropic(model="test")
     message = HumanMessage(content="Hello")
     response = chat([message])
     assert isinstance(response, AIMessage)
@@ -26,7 +26,7 @@ def test_anthropic_call() -> None:
 
 def test_anthropic_streaming() -> None:
     """Test streaming tokens from anthropic."""
-    chat = ChatAnthropic(model="bare-nano-0", streaming=True)
+    chat = ChatAnthropic(model="test", streaming=True)
     message = HumanMessage(content="Hello")
     response = chat([message])
     assert isinstance(response, AIMessage)
@@ -38,11 +38,12 @@ def test_anthropic_streaming_callback() -> None:
     callback_handler = FakeCallbackHandler()
     callback_manager = CallbackManager([callback_handler])
     chat = ChatAnthropic(
+        model="test",
         streaming=True,
         callback_manager=callback_manager,
         verbose=True,
     )
-    message = HumanMessage(content="Write me a sentence with 100 words.")
+    message = HumanMessage(content="Write me a sentence with 10 words.")
     chat([message])
     assert callback_handler.llm_streams > 1
 
@@ -53,6 +54,7 @@ async def test_anthropic_async_streaming_callback() -> None:
     callback_handler = FakeCallbackHandler()
     callback_manager = CallbackManager([callback_handler])
     chat = ChatAnthropic(
+        model="test",
         streaming=True,
         callback_manager=callback_manager,
         verbose=True,