pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-08 20:42:30 +08:00
Code Issues Packages Projects Releases Wiki Activity

Format file format from Windows/dos to Unix (#1949)

Browse Source 
### What problem does this PR solve?

Related source file is in Windows/DOS format, they are format to Unix
format.

### Type of change

- [x] Refactoring

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
This commit is contained in:
Jin Hai
2024-08-15 09:17:36 +08:00
committed by GitHub
parent 1328d715db
commit 6b3a40be5c
108 changed files with 36399 additions and 36399 deletions
Download Patch File Download Diff File Expand all files Collapse all files

124
agent/test/dsl_examples/keyword_wikipedia_and_generate.json
View File

@ -1,62 +1,62 @@
{
"components": {
"begin": {
"obj":{
"component_name": "Begin",
"params": {
"prologue": "Hi there!"
}
},
"downstream": ["answer:0"],
"upstream": []
},
"answer:0": {
"obj": {
"component_name": "Answer",
"params": {}
},
"downstream": ["keyword:0"],
"upstream": ["begin"]
},
"keyword:0": {
"obj": {
"component_name": "KeywordExtract",
"params": {
"llm_id": "deepseek-chat",
"prompt": "- Role: You're a question analyzer.\n - Requirements:\n - Summarize user's question, and give top %s important keyword/phrase.\n - Use comma as a delimiter to separate keywords/phrases.\n - Answer format: (in language of user's question)\n - keyword: ",
"temperature": 0.2,
"top_n": 1
}
},
"downstream": ["wikipedia:0"],
"upstream": ["answer:0"]
},
"wikipedia:0": {
"obj":{
"component_name": "Wikipedia",
"params": {
"top_n": 10
}
},
"downstream": ["generate:0"],
"upstream": ["keyword:0"]
},
"generate:1": {
"obj": {
"component_name": "Generate",
"params": {
"llm_id": "deepseek-chat",
"prompt": "You are an intelligent assistant. Please answer the question based on content from Wikipedia. When the answer from Wikipedia is incomplete, you need to output the URL link of the corresponding content as well. When all the content searched from Wikipedia is irrelevant to the question, your answer must include the sentence, \"The answer you are looking for is not found in the Wikipedia!\". Answers need to consider chat history.\n The content of Wikipedia is as follows:\n {input}\n The above is the content of Wikipedia.",
"temperature": 0.2
}
},
"downstream": ["answer:0"],
"upstream": ["wikipedia:0"]
}
},
"history": [],
"path": [],
"messages": [],
"reference": {},
"answer": []
}
{
"components": {
"begin": {
"obj":{
"component_name": "Begin",
"params": {
"prologue": "Hi there!"
}
},
"downstream": ["answer:0"],
"upstream": []
},
"answer:0": {
"obj": {
"component_name": "Answer",
"params": {}
},
"downstream": ["keyword:0"],
"upstream": ["begin"]
},
"keyword:0": {
"obj": {
"component_name": "KeywordExtract",
"params": {
"llm_id": "deepseek-chat",
"prompt": "- Role: You're a question analyzer.\n - Requirements:\n - Summarize user's question, and give top %s important keyword/phrase.\n - Use comma as a delimiter to separate keywords/phrases.\n - Answer format: (in language of user's question)\n - keyword: ",
"temperature": 0.2,
"top_n": 1
}
},
"downstream": ["wikipedia:0"],
"upstream": ["answer:0"]
},
"wikipedia:0": {
"obj":{
"component_name": "Wikipedia",
"params": {
"top_n": 10
}
},
"downstream": ["generate:0"],
"upstream": ["keyword:0"]
},
"generate:1": {
"obj": {
"component_name": "Generate",
"params": {
"llm_id": "deepseek-chat",
"prompt": "You are an intelligent assistant. Please answer the question based on content from Wikipedia. When the answer from Wikipedia is incomplete, you need to output the URL link of the corresponding content as well. When all the content searched from Wikipedia is irrelevant to the question, your answer must include the sentence, \"The answer you are looking for is not found in the Wikipedia!\". Answers need to consider chat history.\n The content of Wikipedia is as follows:\n {input}\n The above is the content of Wikipedia.",
"temperature": 0.2
}
},
"downstream": ["answer:0"],
"upstream": ["wikipedia:0"]
}
},
"history": [],
"path": [],
"messages": [],
"reference": {},
"answer": []
}