api/python_api_reference.md

DRAFT Python API Reference

THE API REFERENCES BELOW ARE STILL UNDER DEVELOPMENT.

:::tip NOTE Knowledgebase APIs :::

Create knowledge base

RAGFlow.create_dataset(
    name: str,
    avatar: str = "",
    description: str = "",
    language: str = "English",
    permission: str = "me", 
    document_count: int = 0,
    chunk_count: int = 0,
    parse_method: str = "naive",
    parser_config: DataSet.ParserConfig = None
) -> DataSet

Creates a knowledge base (dataset).

Parameters

name: str, Required

The unique name of the dataset to create. It must adhere to the following requirements:

• Permitted characters include:
  • English letters (a-z, A-Z)
  • Digits (0-9)
  • "_" (underscore)
• Must begin with an English letter or underscore.
• Maximum 65,535 characters.
• Case-insensitive.

avatar: str

Base64 encoding of the avatar. Defaults to ""

description

tenant_id: str

The id of the tenant associated with the created dataset is used to identify different users. Defaults to None.

• If creating a dataset, tenant_id must not be provided.
• If updating a dataset, tenant_id can't be changed.

description: str

The description of the created dataset. Defaults to "".

language: str

The language setting of the created dataset. Defaults to "English". ????????????

permission

Specify who can operate on the dataset. Defaults to "me".

document_count: int

The number of documents associated with the dataset. Defaults to 0.

chunk_count: int

The number of data chunks generated or processed by the created dataset. Defaults to 0.

parse_method, str

The method used by the dataset to parse and process data. Defaults to "naive".

parser_config

The parser configuration of the dataset. A ParserConfig object contains the following attributes:

• chunk_token_count: Defaults to 128.
• layout_recognize: Defaults to True.
• delimiter: Defaults to '\n!?。；！？'.
• task_page_size: Defaults to 12.

Returns

• Success: A dataset object.
• Failure: Exception

Examples

from ragflow import RAGFlow

rag_object = RAGFlow(api_key="<YOUR_API_KEY>", base_url="http://<YOUR_BASE_URL>:9380")
ds = rag_object.create_dataset(name="kb_1")

---

Delete knowledge bases

RAGFlow.delete_datasets(ids: list[str] = None)

Deletes knowledge bases by name or ID.

Parameters

ids

The IDs of the knowledge bases to delete.

Returns

• Success: No value is returned.
• Failure: Exception

Examples

rag.delete_datasets(ids=["id_1","id_2"])

---

List knowledge bases

RAGFlow.list_datasets(
    page: int = 1, 
    page_size: int = 1024, 
    orderby: str = "create_time", 
    desc: bool = True,
    id: str = None,
    name: str = None
) -> List[DataSet]

Retrieves a list of knowledge bases.

Parameters

page: int

The current page number to retrieve from the paginated results. Defaults to 1.

page_size: int

The number of records on each page. Defaults to 1024.

order_by: str

The field by which the records should be sorted. This specifies the attribute or column used to order the results. Defaults to "create_time".

desc: bool

Whether the sorting should be in descending order. Defaults to True.

id: str

The id of the dataset to be got. Defaults to None.

name: str

The name of the dataset to be got. Defaults to None.

Returns

• Success: A list of DataSet objects representing the retrieved knowledge bases.
• Failure: Exception.

Examples

List all knowledge bases

for ds in rag_object.list_datasets():
    print(ds)

Retrieve a knowledge base by ID

dataset = rag_object.list_datasets(id = "id_1")
print(dataset[0])

---

Update knowledge base

DataSet.update(update_message: dict)

Updates the current knowledge base.

Parameters

update_message: dict[str, str|int], Required

• "name": str The name of the knowledge base to update.
• "tenant_id": str The "tenant_id you get after calling create_dataset().
• "embedding_model": str The embedding model for generating vector embeddings.
  • Ensure that "chunk_count" is 0 before updating "embedding_model".
• "parser_method": str
  • "naive": General
  • "manual: Manual
  • "qa": Q&A
  • "table": Table
  • "paper": Paper
  • "book": Book
  • "laws": Laws
  • "presentation": Presentation
  • "picture": Picture
  • "one":One
  • "knowledge_graph": Knowledge Graph
  • "email": Email

Returns

• Success: No value is returned.
• Failure: Exception

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="<YOUR_API_KEY>", base_url="http://<YOUR_BASE_URL>:9380")
ds = rag.list_datasets(name="kb_1")
ds.update({"embedding_model":"BAAI/bge-zh-v1.5", "parse_method":"manual"})

---

:::tip API GROUPING File management inside knowledge base :::

Upload document

RAGFLOW.upload_document(ds:DataSet, name:str, blob:bytes)-> bool

Parameters

name

blob

Returns

Examples

---

Retrieve document

RAGFlow.get_document(id:str=None,name:str=None) -> Document

Parameters

id: str, Required

ID of the document to retrieve.

name: str

Name or title of the document.

Returns

A document object containing the following attributes:

id: str

Id of the retrieved document. Defaults to "".

thumbnail: str

Thumbnail image of the retrieved document. Defaults to "".

knowledgebase_id: str

Knowledge base ID related to the document. Defaults to "".

parser_method: str

Method used to parse the document. Defaults to "".

parser_config: ParserConfig

Configuration object for the parser. Defaults to None.

source_type: str

Source type of the document. Defaults to "".

type: str

Type or category of the document. Defaults to "".

created_by: str

Creator of the document. Defaults to "".

name: str

string '' Name or title of the document. Defaults to "".

size: int

Size of the document in bytes or some other unit. Defaults to 0.

token_count: int

Number of tokens in the document. Defaults to "".

chunk_count: int

Number of chunks the document is split into. Defaults to 0.

progress: float

Current processing progress as a percentage. Defaults to 0.0.

progress_msg: str

Message indicating current progress status. Defaults to "".

process_begin_at: datetime

Start time of the document processing. Defaults to None.

process_duation: float

Duration of the processing in seconds or minutes. Defaults to 0.0.

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
doc = rag.get_document(id="wdfxb5t547d",name='testdocument.txt')
print(doc)

---

Save document settings

Document.save() -> bool

Returns

bool

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
doc = rag.get_document(id="wdfxb5t547d")
doc.parser_method= "manual"
doc.save()

---

Download document

Document.download() -> bytes

Returns

bytes of the document.

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
doc = rag.get_document(id="wdfxb5t547d")
open("~/ragflow.txt", "w+").write(doc.download())
print(doc) 

---

List documents

Dataset.list_docs(keywords: str=None, offset: int=0, limit:int = -1) -> List[Document]

Parameters

keywords: str

List documents whose name has the given keywords. Defaults to None.

offset: int

The beginning number of records for paging. Defaults to 0.

limit: int

Records number to return, -1 means all of them. Records number to return, -1 means all of them.

Returns

List[Document]

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
ds = rag.create_dataset(name="kb_1")

filename1 = "~/ragflow.txt"
rag.create_document(ds, name=filename1 , blob=open(filename1 , "rb").read())

filename2 = "~/infinity.txt"
rag.create_document(ds, name=filename2 , blob=open(filename2 , "rb").read())

for d in ds.list_docs(keywords="rag", offset=0, limit=12):
    print(d)

---

Delete documents

Document.delete() -> bool

Returns

bool description: delete success or not

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
ds = rag.create_dataset(name="kb_1")

filename1 = "~/ragflow.txt"
rag.create_document(ds, name=filename1 , blob=open(filename1 , "rb").read())

filename2 = "~/infinity.txt"
rag.create_document(ds, name=filename2 , blob=open(filename2 , "rb").read())
for d in ds.list_docs(keywords="rag", offset=0, limit=12):
    d.delete()

---

Parse document

Document.async_parse() -> None
RAGFLOW.async_parse_documents() -> None

Parameters

????????????????????????????????????????????????????

Returns

????????????????????????????????????????????????????

Examples

#document parse and cancel
rag = RAGFlow(API_KEY, HOST_ADDRESS)
ds = rag.create_dataset(name="dataset_name")
name3 = 'ai.pdf'
path = 'test_data/ai.pdf'
rag.create_document(ds, name=name3, blob=open(path, "rb").read())
doc = rag.get_document(name="ai.pdf")
doc.async_parse()
print("Async parsing initiated")

---

Cancel document parsing

rag.async_cancel_parse_documents(ids)
RAGFLOW.async_cancel_parse_documents()-> None

Parameters

ids, list[]

Returns

?????????????????????????????????????????????????

Examples

#documents parse and cancel
rag = RAGFlow(API_KEY, HOST_ADDRESS)
ds = rag.create_dataset(name="God5")
documents = [
    {'name': 'test1.txt', 'path': 'test_data/test1.txt'},
    {'name': 'test2.txt', 'path': 'test_data/test2.txt'},
    {'name': 'test3.txt', 'path': 'test_data/test3.txt'}
]

# Create documents in bulk
for doc_info in documents:
    with open(doc_info['path'], "rb") as file:
        created_doc = rag.create_document(ds, name=doc_info['name'], blob=file.read())
docs = [rag.get_document(name=doc_info['name']) for doc_info in documents]
ids = [doc.id for doc in docs]

rag.async_parse_documents(ids)
print("Async bulk parsing initiated")

for doc in docs:
    for progress, msg in doc.join(interval=5, timeout=10):
        print(f"{doc.name}: Progress: {progress}, Message: {msg}")

cancel_result = rag.async_cancel_parse_documents(ids)
print("Async bulk parsing cancelled")

---

Join document

??????????????????

Document.join(interval=15, timeout=3600) -> iteral[Tuple[float, str]]

Parameters

interval: int

Time interval in seconds for progress report. Defaults to 15.

timeout: int

Timeout in seconds. Defaults to 3600.

Returns

iteral[Tuple[float, str]]

Add chunk

Document.add_chunk(content:str) -> Chunk

Parameters

content: str, Required

Returns

chunk

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
doc = rag.get_document(id="wdfxb5t547d")
chunk = doc.add_chunk(content="xxxxxxx")

---

Delete chunk

Chunk.delete() -> bool

Returns

bool

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
doc = rag.get_document(id="wdfxb5t547d")
chunk = doc.add_chunk(content="xxxxxxx")
chunk.delete()

---

Save chunk contents

Chunk.save() -> bool

Returns

bool

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
doc = rag.get_document(id="wdfxb5t547d")
chunk = doc.add_chunk(content="xxxxxxx")
chunk.content = "sdfx"
chunk.save()

---

Retrieval

RAGFlow.retrieval(question:str, datasets:List[Dataset], document=List[Document]=None,     offset:int=0, limit:int=6, similarity_threshold:float=0.1, vector_similarity_weight:float=0.3, top_k:int=1024) -> List[Chunk]

Parameters

question: str, Required

The user query or query keywords. Defaults to "".

datasets: List[Dataset], Required

The scope of datasets.

document: List[Document]

The scope of document. None means no limitation. Defaults to None.

offset: int

The beginning point of retrieved records. Defaults to 0.

limit: int

The maximum number of records needed to return. Defaults to 6.

Similarity_threshold: float

The minimum similarity score. Defaults to 0.2.

similarity_threshold_weight: float

The weight of vector cosine similarity, 1 - x is the term similarity weight. Defaults to 0.3.

top_k: int

Number of records engaged in vector cosine computaton. Defaults to 1024.

Returns

List[Chunk]

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
ds = rag.get_dataset(name="ragflow")
name = 'ragflow_test.txt'
path = 'test_data/ragflow_test.txt'
rag.create_document(ds, name=name, blob=open(path, "rb").read())
doc = rag.get_document(name=name)
doc.async_parse()
# Wait for parsing to complete 
for progress, msg in doc.join(interval=5, timeout=30):
    print(progress, msg)
for c in rag.retrieval(question="What's ragflow?", 
             datasets=[ds], documents=[doc], 
             offset=0, limit=6, similarity_threshold=0.1, 
             vector_similarity_weight=0.3,
             top_k=1024
             ):
    print(c)

---

:::tip API GROUPING Chat APIs :::

Create chat

Creates a chat assistant.

RAGFlow.create_chat(
    name: str = "assistant", 
    avatar: str = "path", 
    knowledgebases: List[DataSet] = ["kb1"], 
    llm: Chat.LLM = None, 
    prompt: Chat.Prompt = None
) -> Chat

Returns

• Success: A Chat object representing the chat assistant.
• Failure: Exception

name: str

The name of the chat assistant. Defaults to "assistant".

avatar: str

Base64 encoding of the avatar. Defaults to "".

knowledgebases: list[str]

The associated knowledge bases. Defaults to ["kb1"].

llm: LLM

The llm of the created chat. Defaults to None. When the value is None, a dictionary with the following values will be generated as the default.

• model_name, str
  The chat model name. If it is None, the user's default chat model will be returned.
• temperature, float
  This parameter controls the randomness of predictions by the model. A lower temperature makes the model more confident in its responses, while a higher temperature makes it more creative and diverse. Defaults to 0.1.
• top_p, float
  Also known as “nucleus sampling”, this parameter sets a threshold to select a smaller set of words to sample from. It focuses on the most likely words, cutting off the less probable ones. Defaults to 0.3
• presence_penalty, float
  This discourages the model from repeating the same information by penalizing words that have already appeared in the conversation. Defaults to 0.2.
• frequency penalty, float
  Similar to the presence penalty, this reduces the model’s tendency to repeat the same words frequently. Defaults to 0.7.
• max_token, int
  This sets the maximum length of the model’s output, measured in the number of tokens (words or pieces of words). Defaults to 512.

Prompt: str

Instructions for LLM's responses, including character design, answer length, and language. Defaults to:

You are an intelligent assistant. Please summarize the content of the knowledge base to answer the question. Please list the data in the knowledge base and answer in detail. When all knowledge base content is irrelevant to the question, your answer must include the sentence "The answer you are looking for is not found in the knowledge base!" Answers need to consider chat history.
      Here is the knowledge base:
      {knowledge}
      The above is the knowledge base.

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="<YOUR_API_KEY>", base_url="http://<YOUR_BASE_URL>:9380")
knowledge_base = rag.list_datasets(name="kb_1")
assistant = rag.create_chat("Miss R", knowledgebases=knowledge_base)

---

Update chat

Updates the current chat assistant.

Chat.update(update_message: dict)

Parameters

update_message: dict[str, Any], Required

• "name": str The name of the chat assistant to update.
• "avatar": str Base64 encoding of the avatar. Defaults to ""
• "knowledgebases": list[str] Knowledge bases to update.
• "llm": dict llm settings
  • "model_name", str The chat model name.
  • "temperature", float This parameter controls the randomness of predictions by the model.
  • "top_p", float Also known as “nucleus sampling”, this parameter sets a threshold to select a smaller set of words to sample from.
  • "presence_penalty", float This discourages the model from repeating the same information by penalizing words that have already appeared in the conversation.
  • "frequency penalty", float Similar to the presence penalty, this reduces the model’s tendency to repeat the same words frequently.
  • "max_token", int This sets the maximum length of the model’s output, measured in the number of tokens (words or pieces of words).
• "prompt" : Instructions for LLM's responses, including character design, answer length, and language.

Returns

• Success: No value is returned.
• Failure: Exception

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="<YOUR_API_KEY>", base_url="http://<YOUR_BASE_URL>:9380")
knowledge_base = rag.list_datasets(name="kb_1")
assistant = rag.create_chat("Miss R", knowledgebases=knowledge_base)
assistant.update({"llm": {"temperature":0.8}})

---

Delete chats

Deletes specified chat assistants.

RAGFlow.delete_chats(ids: List[str] = None)

Parameters

ids

IDs of the chat assistants to delete.

Returns

• Success: No value is returned.
• Failure: Exception

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="<YOUR_API_KEY>", base_url="http://<YOUR_BASE_URL>:9380")
rag.delete_chats(ids=["id_1","id_2"])

---

List chats

RAGFlow.list_chats(
    page: int = 1, 
    page_size: int = 1024, 
    orderby: str = "create_time", 
    desc: bool = True,
    id: str = None,
    name: str = None
) -> List[Chat]

Parameters

page

The current page number to retrieve from the paginated results. Defaults to 1.

page_size

The number of records on each page. Defaults to 1024.

order_by

The attribute by which the results are sorted. Defaults to "create_time".

desc

Indicates whether to sort the results in descending order. Defaults to True.

id: string

The ID of the chat to be retrieved. Defaults to None.

name: string

The name of the chat to be retrieved. Defaults to None.

Returns

• Success: A list of Chat objects representing the retrieved knowledge bases.
• Failure: Exception.

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="<YOUR_API_KEY>", base_url="http://<YOUR_BASE_URL>:9380")
for assistant in rag.list_chats():
    print(assistant)

---

:::tip API GROUPING Chat-session APIs :::

Create session

Chat.create_session(name: str = "New session") -> Session

Returns

A session object.

id: str

The id of the created session is used to identify different sessions.

• id can not be provided in creating

name: str

The name of the created session. Defaults to "New session".

messages: List[Message]

The messages of the created session.

• messages cannot be provided.

Defaults:

??????????????????????????????????????????????????????????????????????????????????????????????

[{"role": "assistant", "content": "Hi! I am your assistant，can I help you?"}]

chat_id: str

The id of associated chat

• chat_id can't be changed

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
assi = rag.list_chats(name="Miss R")
assi = assi[0]
sess = assi.create_session()

Update session

Session.update(update_message:dict)

Returns

no return

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
assi = rag.list_chats(name="Miss R")
assi = assi[0]
sess = assi.create_session("new_session")
sess.update({"name": "Updated session"...})

---

Chat

Session.ask(question: str, stream: bool = False) -> Optional[Message, iter[Message]]

Parameters

question: str, Required

The question to start an AI chat. Defaults to None. ???????????????????

stream: bool

The approach of streaming text generation. When stream is True, it outputs results in a streaming fashion; otherwise, it outputs the complete result after the model has finished generating.

Returns

[Message, iter[Message]]

id: str

The id of the message. id is automatically generated. Defaults to None. ???????????????????

content: str

The content of the message. Defaults to "Hi! I am your assistant, can I help you?".

reference: List[Chunk]

The auto-generated reference of the message. Each chunk object includes the following attributes:

• id: str
  The id of the chunk. ?????????????????
• content: str
  The content of the chunk. Defaults to None. ?????????????????????
• document_id: str
  The ID of the document being referenced. Defaults to "".
• document_name: str
  The name of the referenced document being referenced. Defaults to "".
• knowledgebase_id: str
  The id of the knowledge base to which the relevant document belongs. Defaults to "".
• image_id: str
  The id of the image related to the chunk. Defaults to "".
• similarity: float A general similarity score, usually a composite score derived from various similarity measures . This score represents the degree of similarity between two objects. The value ranges between 0 and 1, where a value closer to 1 indicates higher similarity. Defaults to None. ????????????????????????????????????
• vector_similarity: float
  A similarity score based on vector representations. This score is obtained by converting texts, words, or objects into vectors and then calculating the cosine similarity or other distance measures between these vectors to determine the similarity in vector space. A higher value indicates greater similarity in the vector space. Defaults to None. ?????????????????????????????????
• term_similarity: float
  The similarity score based on terms or keywords. This score is calculated by comparing the similarity of key terms between texts or datasets, typically measuring how similar two words or phrases are in meaning or context. A higher value indicates a stronger similarity between terms. Defaults to None. ???????????????????
• position: List[string]
  Indicates the position or index of keywords or specific terms within the text. An array is typically used to mark the location of keywords or specific elements, facilitating precise operations or analysis of the text. Defaults to None. ??????????????

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
assi = rag.list_chats(name="Miss R")
assi = assi[0]
sess = assi.create_session()    

print("\n==================== Miss R =====================\n")
print(assi.get_prologue())

while True:
    question = input("\n==================== User =====================\n> ")
    print("\n==================== Miss R =====================\n")
    
    cont = ""
    for ans in sess.ask(question, stream=True):
        print(ans.content[len(cont):], end='', flush=True)
        cont = ans.content

---

List sessions

Chat.list_sessions(
    page: int = 1, 
    page_size: int = 1024, 
    orderby: str = "create_time", 
    desc: bool = True,
    id: str = None,
    name: str = None
) -> List[Session]

Returns

List[Session] description: the List contains information about multiple assistant object, with each dictionary containing information about one assistant.

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
assi = rag.list_chats(name="Miss R")
assi = assi[0]
for sess in assi.list_sessions():
    print(sess)

Parameters

page: int

The current page number to retrieve from the paginated data. This parameter determines which set of records will be fetched.

• 1

page_size: int

The number of records to retrieve per page. This controls how many records will be included in each page.

• 1024

orderby: string

The field by which the records should be sorted. This specifies the attribute or column used to order the results.

• "create_time"

desc: bool

A boolean flag indicating whether the sorting should be in descending order.

• True

id: string

The ID of the chat to be retrieved.

• None

name: string

The name of the chat to be retrieved.
- None

Delete session

Chat.delete_sessions(ids:List[str] = None)

Returns

no return

Examples

from ragflow import RAGFlow

rag = RAGFlow(api_key="xxxxxx", base_url="http://xxx.xx.xx.xxx:9380")
assi = rag.list_chats(name="Miss R")
assi = assi[0]
assi.delete_sessions(ids=["id_1","id_2"])

Parameters

ids: List[string]

IDs of the sessions to be deleted.

• None