"description":"A chat assistant template that integrates information extracted from a knowledge base and web searches to respond to queries. Let's begin by setting up your knowledge base in 'Retrieval'!",
"prompt":"Role: You are an intelligent assistant. \nTask: Chat with user. Answer the question based on the provided content from: Knowledge Base, Wikipedia, Duckduckgo, Baidu.\nRequirements:\n - Answer should be in markdown format.\n - Answer should include all sources(Knowledge Base, Wikipedia, Duckduckgo, Baidu) as long as they are relevant, and label the sources of the cited content separately.\n - Attach URL links to the content which is quoted from Wikipedia, DuckDuckGo or Baidu.\n - Do not make thing up when there's no relevant information to user's question. \n\n## Knowledge base content\n {kb_input}\n\n\n## Wikipedia content\n{wikipedia}\n\n\n## Duckduckgo content\n{duckduckgo}\n\n\n## Baidu content\n{baidu}",
"empty_response":"The answer you want was not found in the knowledge base!",
"kb_ids":[],
"keywords_similarity_weight":0.3,
"similarity_threshold":0.2,
"top_n":8
},
"label":"Retrieval",
"name":"Search KB"
},
"dragging":false,
"height":46,
"id":"Retrieval:SilentCamelsStick",
"position":{
"x":-641.3113750640641,
"y":-4.669746081545384
},
"positionAbsolute":{
"x":-641.3113750640641,
"y":-4.669746081545384
},
"selected":true,
"sourcePosition":"right",
"targetPosition":"left",
"type":"retrievalNode",
"width":200
},
{
"data":{
"form":{
"cite":true,
"frequencyPenaltyEnabled":true,
"frequency_penalty":0.7,
"llm_id":"deepseek-chat@DeepSeek",
"maxTokensEnabled":true,
"max_tokens":256,
"message_history_window_size":12,
"parameter":"Precise",
"parameters":[
{
"component_id":"Retrieval:SilentCamelsStick",
"id":"2a7a72d0-29da-4cfc-8745-57d5ee2251c3",
"key":"kb_input"
},
{
"component_id":"DuckDuckGo:SoftButtonsRefuse",
"id":"5f7239b7-aa50-431d-a8b7-871cf355b3ca",
"key":"duckduckgo"
},
{
"component_id":"Wikipedia:WittyRiceLearn",
"id":"e6f8b466-d857-482b-b068-e5dcc593ee79",
"key":"wikipedia"
},
{
"component_id":"Baidu:OliveAreasCall",
"id":"d0e07768-bdad-4eee-beff-93a68ec83dd2",
"key":"baidu"
}
],
"presencePenaltyEnabled":true,
"presence_penalty":0.4,
"prompt":"Role: You are an intelligent assistant. \nTask: Chat with user. Answer the question based on the provided content from: Knowledge Base, Wikipedia, Duckduckgo, Baidu.\nRequirements:\n - Answer should be in markdown format.\n - Answer should include all sources(Knowledge Base, Wikipedia, Duckduckgo, Baidu) as long as they are relevant, and label the sources of the cited content separately.\n - Attach URL links to the content which is quoted from Wikipedia, DuckDuckGo or Baidu.\n - Do not make thing up when there's no relevant information to user's question. \n\n## Knowledge base content\n {kb_input}\n\n\n## Wikipedia content\n{wikipedia}\n\n\n## Duckduckgo content\n{duckduckgo}\n\n\n## Baidu content\n{baidu}",
"temperature":0.1,
"temperatureEnabled":true,
"topPEnabled":true,
"top_p":0.3
},
"label":"Generate",
"name":"LLM"
},
"dragging":false,
"height":254,
"id":"Generate:GreatCarrotsDecide",
"position":{
"x":-633.3943275994267,
"y":141.79563201424583
},
"positionAbsolute":{
"x":-633.3943275994267,
"y":141.79563201424583
},
"selected":false,
"sourcePosition":"right",
"targetPosition":"left",
"type":"generateNode",
"width":200
},
{
"data":{
"form":{
"text":"The large model answers the user's query based on the content retrieved from different search engines and knowledge bases, returning an answer to the user's question."
},
"label":"Note",
"name":"N: LLM"
},
"dragHandle":".note-drag-handle",
"dragging":false,
"height":144,
"id":"Note:CuteSchoolsWear",
"position":{
"x":-628.5256394373041,
"y":412.60472782016245
},
"positionAbsolute":{
"x":-628.5256394373041,
"y":412.60472782016245
},
"resizing":false,
"selected":false,
"sourcePosition":"right",
"style":{
"height":144,
"width":443
},
"targetPosition":"left",
"type":"noteNode",
"width":443
},
{
"data":{
"form":{
"text":"Complete questions by conversation history.\nUser: What's RAGFlow?\nAssistant: RAGFlow is xxx.\nUser: How to deloy it?\n\nRefine it: How to deploy RAGFlow?"
},
"label":"Note",
"name":"N: Refine question"
},
"dragging":false,
"height":209,
"id":"Note:CuteRavensBehave",
"position":{
"x":-921.2271023677847,
"y":-381.3182401779728
},
"positionAbsolute":{
"x":-921.2271023677847,
"y":-381.3182401779728
},
"resizing":false,
"selected":false,
"sourcePosition":"right",
"style":{
"height":209,
"width":266
},
"targetPosition":"left",
"type":"noteNode",
"width":266,
"dragHandle":".note-drag-handle"
},
{
"data":{
"form":{
"text":"Based on the user's question, searches the knowledge base and returns the retrieved content."
},
"label":"Note",
"name":"N: Search KB"
},
"dragging":false,
"height":128,
"id":"Note:RudeRulesLeave",
"position":{
"x":-917.896611693436,
"y":-3.570404025438563
},
"positionAbsolute":{
"x":-917.896611693436,
"y":-3.570404025438563
},
"selected":false,
"sourcePosition":"right",
"targetPosition":"left",
"type":"noteNode",
"width":269,
"dragHandle":".note-drag-handle"
},
{
"data":{
"form":{
"text":"Based on the keywords, searches on Wikipedia and returns the found content."
},
"label":"Note",
"name":"N: Wiukipedia"
},
"dragging":false,
"height":128,
"id":"Note:DryActorsTry",
"position":{
"x":49.68127281474659,
"y":-16.899164744846445
},
"positionAbsolute":{
"x":49.68127281474659,
"y":-16.899164744846445
},
"resizing":false,
"selected":false,
"sourcePosition":"right",
"style":{
"height":128,
"width":281
},
"targetPosition":"left",
"type":"noteNode",
"width":281,
"dragHandle":".note-drag-handle"
},
{
"data":{
"form":{
"text":"Based on the keywords, searches on Baidu and returns the found content."
},
"label":"Note",
"name":"N :Baidu"
},
"dragging":false,
"height":128,
"id":"Note:HonestShirtsNail",
"position":{
"x":43.964372149616565,
"y":-151.26282396084338
},
"positionAbsolute":{
"x":43.964372149616565,
"y":-151.26282396084338
},
"selected":false,
"sourcePosition":"right",
"targetPosition":"left",
"type":"noteNode",
"width":269,
"dragHandle":".note-drag-handle"
},
{
"data":{
"form":{
"text":"Based on the keywords, searches on DuckDuckGo and returns the found content."
},
"label":"Note",
"name":"N: DuckduckGo"
},
"dragging":false,
"height":145,
"id":"Note:OddBreadsFix",
"position":{
"x":-237.54626926201882,
"y":-381.56637252684175
},
"positionAbsolute":{
"x":-237.54626926201882,
"y":-381.56637252684175
},
"resizing":false,
"selected":false,
"sourcePosition":"right",
"style":{
"height":145,
"width":201
},
"targetPosition":"left",
"type":"noteNode",
"width":201,
"dragHandle":".note-drag-handle"
},
{
"data":{
"form":{
"text":"The large model generates keywords based on the user's question for better retrieval."
