docs: Add more detail to some template notes (#9230)
* update travel agent template description * update vector store rag description * update sequential task agent description * update chat memory template description * update financial report parser description * update basic prompting template description --------- Co-authored-by: Edwin Jose <edwin.jose@datastax.com>
This commit is contained in:
April I. Murphy
GitHub
parent
15637403ad
commit
3369e28f3b
6 changed files with 6 additions and 6 deletions
|
|
@ -516,7 +516,7 @@
|
|||
"data": {
|
||||
"id": "undefined-qkI8p",
|
||||
"node": {
|
||||
"description": "## 📖 README\n\nPerform basic prompting with a Language model component.\n\n#### Quick Start\n- Add your **OpenAI API key** to the **Language Model** component.\n- Open the **Playground** to chat with your bot.\n\n#### Next steps:\nExperiment by changing the prompt and the Language model temperature to see how the bot's responses change.",
|
||||
"description": "This template demonstrates a standard chat flow with additional instructions provided by a prompt. Prompts provide instructions and inputs for a Large Language Model (LLM) beyond the standard user-provided chat input. In this example, the prompt describes the LLM's role and persona.\n\n## Quick Start\n1. Add your **OpenAI API Key** to the **Language Model** component, or select a different provider and model.\n2. Open the **Playground** to start the chat and run the flow.\n\n## Next steps\nChange the prompt template, model, or model settings, such as **Temperature**, and then see how the responses change with these different inputs.\n💡 Some component settings are hidden by default; to view all settings click **Controls** in each component's header menu.\n💡 You can use curly braces to create variables in your template, such as `{variable}`. These can be populated from other components, with Langflow global variables, or at runtime.",
|
||||
"display_name": "Read Me",
|
||||
"documentation": "",
|
||||
"template": {
|
||||
|
|
|
|||
|
|
@ -780,7 +780,7 @@
|
|||
"data": {
|
||||
"id": "note-2715s",
|
||||
"node": {
|
||||
"description": "\n# Financial Report Parser\n\nThis template extracts key financial metrics from a given financial report text using OpenAI's GPT-4o-mini model. The extracted data is structured and formatted for chat consumption.\n\n## Prerequisites\n\n- **[OpenAI API Key](https://platform.openai.com/)**\n\n## Quickstart\n\n1. Add your OpenAI API key to the Language model.\n2. To run the flow, click **Playground**.\nThe **Chat Input** component in this template is pre-loaded with a sample financial report for demonstrating how structured data is extracted.\n\n* The **Language Model** model component identifies and retrieves Gross Profit, EBITDA, Net Income, and Operating Expenses from the financial report.\n* The **Structured Output** component formats extracted data into a structured format for better readability and further processing.\n* The **Parser** component converts extracted data into formatted messages for chat consumption.\n\n\n\n\n\n",
|
||||
"description": "This template extracts key financial metrics from a given financial report text. The extracted data is structured and formatted for chat consumption.\n\n## Quickstart\n\n1. Add your OpenAI API key to the **Language Model** component, or select a different provider and model.\n2. Open the **Playground** to start the chat and run the flow.\nFor this example, the **Chat Input** component is pre-loaded with a sample financial report. The **Language Model** component identifies and retrieves the gross profit, EBITDA, net income, and operating expenses information from the financial report. Then, the **Structured Output** component formats extracted data into a structured format for better readability and further processing. Finally, the **Parser** component converts extracted data into a messages to be returned to the user.",
|
||||
"display_name": "",
|
||||
"documentation": "",
|
||||
"template": {}
|
||||
|
|
|
|||
|
|
@ -736,7 +736,7 @@
|
|||
"data": {
|
||||
"id": "note-hLqzd",
|
||||
"node": {
|
||||
"description": "# Memory Chatbot\n\nThis flow extends the basic prompting flow with a Message history component that stores up to 100 previous chat messages and uses them to provide context for the current conversation.\n\n## Prerequisites\n\n* [OpenAI API Key](https://platform.openai.com/)\n\n## Quickstart\n\n1. In the **Language Model** component, add your OpenAI API Key.\n\n2. Open the **Playground**. Tell the chat your name.\n\n3. Start a new chat session in the Playground, and ask, `what is my name`. The Prompt component will still remember your name, because it's connected to the Message History component.\n\n",
|
||||
"description": "This flow extends the **Basic Prompting** template by adding a **Message History** component that can retrieve up to 100 previous chat messages as context for the current conversation.\n\n## Quickstart\n1. In the **Language Model** component, add your OpenAI API key, or select a different provider and model.\n2. Open the **Playground**, and then tell the LLM your name.\n3. Start a new chat session in the Playground, and then ask, `what is my name`. The LLM is able to retrieve your name from the stored chat history.\n\n## About the Message History component\nThe **Language Model** and **Agent** components have built-in chat memory that is enabled by default and functionally the same as the **Message History** component.\nOnly use the **Message History** component when you want to store or retrieve chat memory from an external chat memory database, or when you need to retrieve chat memory outside of the current session context, such as in a non-chat flow or by supplying memories from other chats to a different session. For more information, see [Store chat memory](https://docs.langflow.org/memory#store-chat-memory).",
|
||||
"display_name": "",
|
||||
"documentation": "",
|
||||
"template": {}
|
||||
|
|
|
|||
|
|
@ -2191,7 +2191,7 @@
|
|||
"data": {
|
||||
"id": "note-VLQWH",
|
||||
"node": {
|
||||
"description": "# 📖 README\n## Overview\nThis flow demonstrates how to chain multiple AI agents for comprehensive research and analysis. Each agent specializes in different aspects of the research process, building upon the previous agent's work.\n\n## Quickstart\n- Add your **OpenAI API Key** to the **Researcher Agent**, **Finance Agent**, and **Analysis & Editor Agent** or change the provider and provide the neccesary credentials\n- Add your **Tavily API Key** to the **Tavily AI Search**\n\n## Using the Flow\n\n**Input Your Query** 🎯\n - Be specific and clear\n - Include key aspects you want analyzed\n - Example:\n ```\n Good: \"Should I invest in Tesla (TSLA)? Focus on AI development impact\"\n Bad: \"Tell me about Tesla\"\n ```\n\n*Note: This flow template uses financial analysis as an example but can be adapted for any research-intensive task requiring multiple perspectives and data sources.*",
|
||||
"description": "This flow demonstrates how to chain multiple AI agents for comprehensive research and analysis. Each agent specializes in different aspects of the research process, building upon the previous agent's work. \n\n## Quickstart\n1. Add your **OpenAI API Key** to the three **Agent** components. If you want to use a different model or provider, change the **Model Provider**, **Model Name**, and **API Key** fields accordingly.\n2. Add your **Tavily API Key** to the **Tavily AI Search** component.\n3. Open the **Playground** and enter a query to run the flow. Be specific, clear, and include key aspects that you want the agents to analyze in a financial perspective.\nBecause this flow includes a financial analysis agent, useful queries should include a financial aspect, such as \"Should I invest in Tesla (TSLA)? Focus on AI development impact\". In contrast, asking the agent, \"Tell me about Tesla\" isn't as useful because it doesn't trigger the financial research agent or provide specific talking points for the other agents to research.\n\n## Next steps\nThis template uses financial analysis as an example. Try adapting it for other research-intensive tasks that require multiple perspectives and data sources.",
|
||||
"display_name": "",
|
||||
"documentation": "",
|
||||
"template": {}
|
||||
|
|
|
|||
|
|
@ -808,7 +808,7 @@
|
|||
"data": {
|
||||
"id": "note-j1Nu2",
|
||||
"node": {
|
||||
"description": "# 📖 README\n\nThe travel planning system is a smart setup that uses several specialized \"agents\" to help plan incredible trips. Imagine each agent as a travel expert focusing on a part of your journey. Here's how it works:\n\n## Quickstart\n1. Add your **OpenAI API Key** to each Agent.\n2. Add your **Search API** key to the Search API component.\n2. Run the flow in the **Playground**.",
|
||||
"description": "# 📖 README\n\nThe travel planning system is a smart setup that uses several specialized Agent components to help plan incredible trips. Imagine each agent as a travel expert focusing on a part of your journey. Here's how it works:\nEach agent has a role defined in the **Agent Instructions** and relevant **Tools** attached. The user submits a query through the **Chat Input**, and the three agents create a complete travel plan based on the user's query.\n\n## Quickstart\n1. Add your **OpenAI API Key** to each Agent.\n2. Add your **Search API** key to the Search API component.\n2. Run the flow in the **Playground**.",
|
||||
"display_name": "",
|
||||
"documentation": "",
|
||||
"template": {}
|
||||
|
|
|
|||
|
|
@ -1018,7 +1018,7 @@
|
|||
"data": {
|
||||
"id": "note-zTYux",
|
||||
"node": {
|
||||
"description": "## 📖 README\n\nLoad your data into a vector database with the 📚 **Load Data** flow, and then use your data as chat context with the 🐕 **Retriever** flow.\n\n**🚨 Add your OpenAI API key as a global variable to easily add it to all of the OpenAI components in this flow.** \n\n**Quick start**\n1. Run the 📚 **Load Data** flow.\n2. Run the 🐕 **Retriever** flow.\n\n**Next steps** \n\n- Experiment by changing the prompt and the loaded data to see how the bot's responses change. \n\nFor more info, see the [Langflow docs](https://docs.langflow.org/starter-projects-vector-store-rag).",
|
||||
"description": "Retrieval Augmented Generation (RAG) is a way of providing additional context to a Large Language Model (LLM) by preloading a vector database with embeddings for relevant content. When a user chats with the LLM, a _similarity search_ retrieves relevant content by comparing an embedding for the user's query against the embeddings in the vector database.\nFor example, a RAG chatbot could be pre-loaded with product data, and then it can help customers find specific products based on their queries.\nThis template has two sub-flows. One flow loads data into your vector store, and the other is the user-driven chat flow that compares a new query against the existing content in your vector database.\n\n## Quickstart\n1. Add your OpenAI API key to the **Language Model** component and the two **Embeddings** components.\n2. Add an Astra application token to the **Astra DB** vector store components, or replace these components with other vector store components available in the **Components** menu.\n**💡 Store your credentials as Langflow global variables 🌐 to simplify token management and reuse in your flows.**\n\n## Run the flows\n1. Load your data into a vector database with the 📚 **Load Data** flow. Select a file to upload in the **File** component, and then click **Play** ▶️ on the **Astra DB** component to run the **Load Data** flow.\n2. Open the **Playground** to start a chat with the 🐕 **Retriever** flow.\n\nOnly the run the **Load Data** flow when you need to populate your vector database with baseline content, such as product data.\nThe **Retriever** flow is the user-facing chat flow. This flow generates an embedding from chat input, runs a similarity search against the vector database to retrieve relevant content, and then passes the original query and the retrieved content to the LLM, which produces the chat response sent to the user.\n\n## Next steps\nExperiment by changing the prompt and the loaded data to see how the LLM's responses change.",
|
||||
"display_name": "Read Me",
|
||||
"documentation": "",
|
||||
"template": {
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue