mirror of
https://github.com/hwchase17/langchain.git
synced 2025-06-20 05:43:55 +00:00
docs: Minor corrections and updates to Cohere docs (#22726)
- **Description:** Update the Cohere's provider and RagRetriever
documentations with latest updates.
- **Twitter handle:** Anirudh1810
This commit is contained in:
Anirudh31415926535
GitHub
parent
40b4a3de6e
commit
4da3d4b18e
@ -2,7 +2,7 @@
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "raw",
|
||||
"id": "afaf8039",
|
||||
"id": "53fbf15f",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"---\n",
|
||||
@ -12,129 +12,103 @@
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "e49f1e0d",
|
||||
"id": "bf733a38-db84-4363-89e2-de6735c37230",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"# ChatCohere\n",
|
||||
"# Cohere\n",
|
||||
"\n",
|
||||
"This doc will help you get started with Cohere [chat models](/docs/concepts/#chat-models). For detailed documentation of all ChatCohere features and configurations head to the [API reference](https://api.python.langchain.com/en/latest/chat_models/langchain_cohere.chat_models.ChatCohere.html).\n",
|
||||
"\n",
|
||||
"For an overview of all Cohere models head to the [Cohere docs](https://docs.cohere.com/docs/models).\n",
|
||||
"\n",
|
||||
"## Overview\n",
|
||||
"### Integration details\n",
|
||||
"\n",
|
||||
"| Class | Package | Local | Serializable | [JS support](https://js.langchain.com/v0.2/docs/integrations/chat/cohere) | Package downloads | Package latest |\n",
|
||||
"| :--- | :--- | :---: | :---: | :---: | :---: | :---: |\n",
|
||||
"| [ChatCohere](https://api.python.langchain.com/en/latest/chat_models/langchain_cohere.chat_models.ChatCohere.html) | [langchain-cohere](https://api.python.langchain.com/en/latest/cohere_api_reference.html) | ❌ | beta | ✅ |  |  |\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",
|
||||
"This notebook covers how to get started with [Cohere chat models](https://cohere.com/chat).\n",
|
||||
"\n",
|
||||
"Head to the [API reference](https://api.python.langchain.com/en/latest/chat_models/langchain_community.chat_models.cohere.ChatCohere.html) for detailed documentation of all attributes and methods."
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "3607d67e-e56c-4102-bbba-df2edc0e109e",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## Setup\n",
|
||||
"\n",
|
||||
"To access Cohere models you'll need to create a Cohere account, get an API key, and install the `langchain-cohere` integration package.\n",
|
||||
"The integration lives in the `langchain-cohere` package. We can install these with:\n",
|
||||
"\n",
|
||||
"### Credentials\n",
|
||||
"```bash\n",
|
||||
"pip install -U langchain-cohere\n",
|
||||
"```\n",
|
||||
"\n",
|
||||
"Head to https://dashboard.cohere.com/welcome/login to sign up to Cohere and generate an API key. Once you've done this set the COHERE_API_KEY environment variable:"
|
||||
"We'll also need to get a [Cohere API key](https://cohere.com/) and set the `COHERE_API_KEY` environment variable:"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "433e8d2b-9519-4b49-b2c4-7ab65b046c94",
|
||||
"execution_count": 11,
|
||||
"id": "2108b517-1e8d-473d-92fa-4f930e8072a7",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"import getpass\n",
|
||||
"import os\n",
|
||||
"\n",
|
||||
"os.environ[\"COHERE_API_KEY\"] = getpass.getpass(\"Enter your Cohere API key: \")"
|
||||
"os.environ[\"COHERE_API_KEY\"] = getpass.getpass()"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "72ee0c4b-9764-423a-9dbf-95129e185210",
|
||||
"id": "cf690fbb",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"If you want to get automated tracing of your model calls you can also set your [LangSmith](https://docs.smith.langchain.com/) API key by uncommenting below:"
|
||||
"It's also helpful (but not needed) to set up [LangSmith](https://smith.langchain.com/) for best-in-class observability"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "a15d341e-3e26-4ca3-830b-5aab30ed66de",
|
||||
"execution_count": 12,
|
||||
"id": "7f11de02",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"# os.environ[\"LANGSMITH_API_KEY\"] = getpass.getpass(\"Enter your LangSmith API key: \")\n",
|
||||
"# os.environ[\"LANGSMITH_TRACING\"] = \"true\""
|
||||
"# os.environ[\"LANGCHAIN_TRACING_V2\"] = \"true\"\n",
|
||||
"# os.environ[\"LANGCHAIN_API_KEY\"] = getpass.getpass()"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "0730d6a1-c893-4840-9817-5e5251676d5d",
|
||||
"id": "4c26754b-b3c9-4d93-8f36-43049bd943bf",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"### Installation\n",
|
||||
"## Usage\n",
|
||||
"\n",
|
||||
"The LangChain Cohere integration lives in the `langchain-cohere` package:"
|
||||
"ChatCohere supports all [ChatModel](/docs/how_to#chat-models) functionality:"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "652d6238-1f87-422a-b135-f5abbb8652fc",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"%pip install -qU langchain-cohere"
|
||||
]
|
||||
"execution_count": 5,
|
||||
"id": "d4a7c55d-b235-4ca4-a579-c90cc9570da9",
|
||||
"metadata": {
|
||||
"tags": []
|
||||
},
|
||||
{
|
||||
"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_cohere import ChatCohere\n",
|
||||
"\n",
|
||||
"llm = ChatCohere(\n",
|
||||
" model=\"command-r-plus\",\n",
|
||||
" temperature=0,\n",
|
||||
" max_tokens=None,\n",
|
||||
" timeout=None,\n",
|
||||
" max_retries=2,\n",
|
||||
" # other params...\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "2b4f3e15",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## Invocation"
|
||||
"from langchain_core.messages import HumanMessage"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 2,
|
||||
"id": "62e0dbc3",
|
||||
"execution_count": 13,
|
||||
"id": "70cf04e8-423a-4ff6-8b09-f11fb711c817",
|
||||
"metadata": {
|
||||
"tags": []
|
||||
},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"chat = ChatCohere()"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 15,
|
||||
"id": "8199ef8f-eb8b-4253-9ea0-6c24a013ca4c",
|
||||
"metadata": {
|
||||
"tags": []
|
||||
},
|
||||
@ -142,110 +116,223 @@
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"AIMessage(content=\"J'adore programmer.\", additional_kwargs={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': 'd84f80f3-4611-46e6-aed0-9d8665a20a11', 'token_count': {'input_tokens': 89, 'output_tokens': 5}}, response_metadata={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': 'd84f80f3-4611-46e6-aed0-9d8665a20a11', 'token_count': {'input_tokens': 89, 'output_tokens': 5}}, id='run-514ab516-ed7e-48ac-b132-2598fb80ebef-0')"
|
||||
"AIMessage(content='4 && 5 \\n6 || 7 \\n\\nWould you like to play a game of odds and evens?', additional_kwargs={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '2076b614-52b3-4082-a259-cc92cd3d9fea', 'token_count': {'prompt_tokens': 68, 'response_tokens': 23, 'total_tokens': 91, 'billed_tokens': 77}}, response_metadata={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '2076b614-52b3-4082-a259-cc92cd3d9fea', 'token_count': {'prompt_tokens': 68, 'response_tokens': 23, 'total_tokens': 91, 'billed_tokens': 77}}, id='run-3475e0c8-c89b-4937-9300-e07d652455e1-0')"
|
||||
]
|
||||
},
|
||||
"execution_count": 2,
|
||||
"execution_count": 15,
|
||||
"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"
|
||||
"messages = [HumanMessage(content=\"1\"), HumanMessage(content=\"2 3\")]\n",
|
||||
"chat.invoke(messages)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 3,
|
||||
"id": "d86145b3-bfef-46e8-b227-4dda5c9c2705",
|
||||
"execution_count": 16,
|
||||
"id": "c5fac0e9-05a4-4fc1-a3b3-e5bbb24b971b",
|
||||
"metadata": {
|
||||
"tags": []
|
||||
},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"AIMessage(content='4 && 5', additional_kwargs={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': 'f0708a92-f874-46ee-9b93-334d616ad92e', 'token_count': {'prompt_tokens': 68, 'response_tokens': 3, 'total_tokens': 71, 'billed_tokens': 57}}, response_metadata={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': 'f0708a92-f874-46ee-9b93-334d616ad92e', 'token_count': {'prompt_tokens': 68, 'response_tokens': 3, 'total_tokens': 71, 'billed_tokens': 57}}, id='run-1635e63e-2994-4e7f-986e-152ddfc95777-0')"
|
||||
]
|
||||
},
|
||||
"execution_count": 16,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"await chat.ainvoke(messages)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 17,
|
||||
"id": "025be980-e50d-4a68-93dc-c9c7b500ce34",
|
||||
"metadata": {
|
||||
"tags": []
|
||||
},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
"J'adore programmer.\n"
|
||||
"4 && 5"
|
||||
]
|
||||
}
|
||||
],
|
||||
"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:"
|
||||
"for chunk in chat.stream(messages):\n",
|
||||
" print(chunk.content, end=\"\", flush=True)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 4,
|
||||
"id": "e197d1d7-a070-4c96-9f8a-a0e86d046e0b",
|
||||
"execution_count": 18,
|
||||
"id": "064288e4-f184-4496-9427-bcf148fa055e",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"AIMessage(content='Ich liebe Programmierung.', additional_kwargs={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '053bebde-4e1d-4d06-8ee6-3446e7afa25e', 'token_count': {'input_tokens': 84, 'output_tokens': 6}}, response_metadata={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '053bebde-4e1d-4d06-8ee6-3446e7afa25e', 'token_count': {'input_tokens': 84, 'output_tokens': 6}}, id='run-53700708-b7fb-417b-af36-1a6fcde38e7d-0')"
|
||||
"[AIMessage(content='4 && 5', additional_kwargs={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '6770ca86-f6c3-4ba3-a285-c4772160612f', 'token_count': {'prompt_tokens': 68, 'response_tokens': 3, 'total_tokens': 71, 'billed_tokens': 57}}, response_metadata={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '6770ca86-f6c3-4ba3-a285-c4772160612f', 'token_count': {'prompt_tokens': 68, 'response_tokens': 3, 'total_tokens': 71, 'billed_tokens': 57}}, id='run-8d6fade2-1b39-4e31-ab23-4be622dd0027-0')]"
|
||||
]
|
||||
},
|
||||
"execution_count": 4,
|
||||
"execution_count": 18,
|
||||
"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",
|
||||
")"
|
||||
"chat.batch([messages])"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "3a5bb5ca-c3ae-4a58-be67-2cd18574b9a3",
|
||||
"id": "f1c56460",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## API reference\n",
|
||||
"## Chaining\n",
|
||||
"\n",
|
||||
"For detailed documentation of all ChatCohere features and configurations head to the API reference: https://api.python.langchain.com/en/latest/chat_models/langchain_cohere.chat_models.ChatCohere.html"
|
||||
"You can also easily combine with a prompt template for easy structuring of user input. We can do this using [LCEL](/docs/concepts#langchain-expression-language-lcel)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 19,
|
||||
"id": "0851b103",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from langchain_core.prompts import ChatPromptTemplate\n",
|
||||
"\n",
|
||||
"prompt = ChatPromptTemplate.from_template(\"Tell me a joke about {topic}\")\n",
|
||||
"chain = prompt | chat"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 20,
|
||||
"id": "ae950c0f-1691-47f1-b609-273033cae707",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"AIMessage(content='What color socks do bears wear?\\n\\nThey don’t wear socks, they have bear feet. \\n\\nHope you laughed! If not, maybe this will help: laughter is the best medicine, and a good sense of humor is infectious!', additional_kwargs={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '6edccf44-9bc8-4139-b30e-13b368f3563c', 'token_count': {'prompt_tokens': 68, 'response_tokens': 51, 'total_tokens': 119, 'billed_tokens': 108}}, response_metadata={'documents': None, 'citations': None, 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '6edccf44-9bc8-4139-b30e-13b368f3563c', 'token_count': {'prompt_tokens': 68, 'response_tokens': 51, 'total_tokens': 119, 'billed_tokens': 108}}, id='run-ef7f9789-0d4d-43bf-a4f7-f2a0e27a5320-0')"
|
||||
]
|
||||
},
|
||||
"execution_count": 20,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"chain.invoke({\"topic\": \"bears\"})"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "12db8d69",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## Tool calling\n",
|
||||
"\n",
|
||||
"Cohere supports tool calling functionalities!"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 7,
|
||||
"id": "337e24af",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from langchain_core.messages import (\n",
|
||||
" HumanMessage,\n",
|
||||
" ToolMessage,\n",
|
||||
")\n",
|
||||
"from langchain_core.tools import tool"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 8,
|
||||
"id": "74d292e7",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"@tool\n",
|
||||
"def magic_function(number: int) -> int:\n",
|
||||
" \"\"\"Applies a magic operation to an integer\n",
|
||||
" Args:\n",
|
||||
" number: Number to have magic operation performed on\n",
|
||||
" \"\"\"\n",
|
||||
" return number + 10\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"def invoke_tools(tool_calls, messages):\n",
|
||||
" for tool_call in tool_calls:\n",
|
||||
" selected_tool = {\"magic_function\": magic_function}[tool_call[\"name\"].lower()]\n",
|
||||
" tool_output = selected_tool.invoke(tool_call[\"args\"])\n",
|
||||
" messages.append(ToolMessage(tool_output, tool_call_id=tool_call[\"id\"]))\n",
|
||||
" return messages\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"tools = [magic_function]"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 9,
|
||||
"id": "ecafcbc6",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"llm_with_tools = chat.bind_tools(tools=tools)\n",
|
||||
"messages = [HumanMessage(content=\"What is the value of magic_function(2)?\")]"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 11,
|
||||
"id": "aa34fc39",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"AIMessage(content='The value of magic_function(2) is 12.', additional_kwargs={'documents': [{'id': 'magic_function:0:2:0', 'output': '12', 'tool_name': 'magic_function'}], 'citations': [ChatCitation(start=34, end=36, text='12', document_ids=['magic_function:0:2:0'])], 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '96a55791-0c58-4e2e-bc2a-8550e137c46d', 'token_count': {'input_tokens': 998, 'output_tokens': 59}}, response_metadata={'documents': [{'id': 'magic_function:0:2:0', 'output': '12', 'tool_name': 'magic_function'}], 'citations': [ChatCitation(start=34, end=36, text='12', document_ids=['magic_function:0:2:0'])], 'search_results': None, 'search_queries': None, 'is_search_required': None, 'generation_id': '96a55791-0c58-4e2e-bc2a-8550e137c46d', 'token_count': {'input_tokens': 998, 'output_tokens': 59}}, id='run-f318a9cf-55c8-44f4-91d1-27cf46c6a465-0')"
|
||||
]
|
||||
},
|
||||
"execution_count": 11,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"res = llm_with_tools.invoke(messages)\n",
|
||||
"while res.tool_calls:\n",
|
||||
" messages.append(res)\n",
|
||||
" messages = invoke_tools(res.tool_calls, messages)\n",
|
||||
" res = llm_with_tools.invoke(messages)\n",
|
||||
"\n",
|
||||
"res"
|
||||
]
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
"kernelspec": {
|
||||
"display_name": "poetry-venv-2",
|
||||
"display_name": "Python 3 (ipykernel)",
|
||||
"language": "python",
|
||||
"name": "poetry-venv-2"
|
||||
"name": "python3"
|
||||
},
|
||||
"language_info": {
|
||||
"codemirror_mode": {
|
||||
@ -257,7 +344,7 @@
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.11.9"
|
||||
"version": "3.9.6"
|
||||
}
|
||||
},
|
||||
"nbformat": 4,
|
||||
|
||||
@ -108,7 +108,7 @@
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"model = Cohere(model=\"command\", max_tokens=256, temperature=0.75)"
|
||||
"model = Cohere(max_tokens=256, temperature=0.75)"
|
||||
]
|
||||
},
|
||||
{
|
||||
|
||||
@ -46,6 +46,55 @@ print(llm.invoke("Come up with a pet name"))
|
||||
```
|
||||
|
||||
Usage of the Cohere (legacy) [LLM model](/docs/integrations/llms/cohere)
|
||||
|
||||
### Tool calling
|
||||
```python
|
||||
from langchain_cohere import ChatCohere
|
||||
from langchain_core.messages import (
|
||||
HumanMessage,
|
||||
ToolMessage,
|
||||
)
|
||||
from langchain_core.tools import tool
|
||||
|
||||
@tool
|
||||
def magic_function(number: int) -> int:
|
||||
"""Applies a magic operation to an integer
|
||||
|
||||
Args:
|
||||
number: Number to have magic operation performed on
|
||||
"""
|
||||
return number + 10
|
||||
|
||||
def invoke_tools(tool_calls, messages):
|
||||
for tool_call in tool_calls:
|
||||
selected_tool = {"magic_function":magic_function}[
|
||||
tool_call["name"].lower()
|
||||
]
|
||||
tool_output = selected_tool.invoke(tool_call["args"])
|
||||
messages.append(ToolMessage(tool_output, tool_call_id=tool_call["id"]))
|
||||
return messages
|
||||
|
||||
tools = [magic_function]
|
||||
|
||||
llm = ChatCohere()
|
||||
llm_with_tools = llm.bind_tools(tools=tools)
|
||||
messages = [
|
||||
HumanMessage(
|
||||
content="What is the value of magic_function(2)?"
|
||||
)
|
||||
]
|
||||
|
||||
res = llm_with_tools.invoke(messages)
|
||||
while res.tool_calls:
|
||||
messages.append(res)
|
||||
messages = invoke_tools(res.tool_calls, messages)
|
||||
res = llm_with_tools.invoke(messages)
|
||||
|
||||
print(res.content)
|
||||
```
|
||||
Tool calling with Cohere LLM can be done by binding the necessary tools to the llm as seen above.
|
||||
An alternative, is to support multi hop tool calling with the ReAct agent as seen below.
|
||||
|
||||
### ReAct Agent
|
||||
|
||||
The agent is based on the paper
|
||||
@ -77,6 +126,7 @@ agent_executor.invoke({
|
||||
"input": "In what year was the company that was founded as Sound of Music added to the S&P 500?",
|
||||
})
|
||||
```
|
||||
The ReAct agent can be used to call multiple tools in sequence.
|
||||
|
||||
### RAG Retriever
|
||||
|
||||
|
||||
@ -34,8 +34,7 @@
|
||||
},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from langchain_cohere import ChatCohere\n",
|
||||
"from langchain_community.retrievers import CohereRagRetriever\n",
|
||||
"from langchain_cohere import ChatCohere, CohereRagRetriever\n",
|
||||
"from langchain_core.documents import Document"
|
||||
]
|
||||
},
|
||||
@ -200,7 +199,7 @@
|
||||
"source": [
|
||||
"docs = rag.invoke(\n",
|
||||
" \"Does langchain support cohere RAG?\",\n",
|
||||
" source_documents=[\n",
|
||||
" documents=[\n",
|
||||
" Document(page_content=\"Langchain supports cohere RAG!\"),\n",
|
||||
" Document(page_content=\"The sky is blue!\"),\n",
|
||||
" ],\n",
|
||||
@ -208,6 +207,14 @@
|
||||
"_pretty_print(docs)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "45a9470f",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"Please note that connectors and documents cannot be used simultaneously. If you choose to provide documents in the `invoke` method, they will take precedence, and connectors will not be utilized for that particular request, as shown in the snippet above!"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
|
||||
@ -8,7 +8,7 @@ import CodeBlock from "@theme-original/CodeBlock";
|
||||
* @typedef {Object} ChatModelTabsProps - Component props.
|
||||
* @property {string} [openaiParams] - Parameters for OpenAI chat model. Defaults to `model="gpt-3.5-turbo-0125"`
|
||||
* @property {string} [anthropicParams] - Parameters for Anthropic chat model. Defaults to `model="claude-3-sonnet-20240229"`
|
||||
* @property {string} [cohereParams] - Parameters for Cohere chat model. Defaults to `model="command-r"`
|
||||
* @property {string} [cohereParams] - Parameters for Cohere chat model. Defaults to `model="command-r-plus"`
|
||||
* @property {string} [fireworksParams] - Parameters for Fireworks chat model. Defaults to `model="accounts/fireworks/models/mixtral-8x7b-instruct"`
|
||||
* @property {string} [groqParams] - Parameters for Groq chat model. Defaults to `model="llama3-8b-8192"`
|
||||
* @property {string} [mistralParams] - Parameters for Mistral chat model. Defaults to `model="mistral-large-latest"`
|
||||
|
||||
@ -111,7 +111,7 @@ class ChatCohere(BaseChatModel, BaseCohere):
|
||||
from langchain_community.chat_models import ChatCohere
|
||||
from langchain_core.messages import HumanMessage
|
||||
|
||||
chat = ChatCohere(model="command", max_tokens=256, temperature=0.75)
|
||||
chat = ChatCohere(max_tokens=256, temperature=0.75)
|
||||
|
||||
messages = [HumanMessage(content="knock knock")]
|
||||
chat.invoke(messages)
|
||||
|
||||
Loading…
Reference in New Issue
Block a user