DashScopeRerank text rerank model-Alibaba Cloud Model Studio(Model Studio)-阿里云帮助中心

DashScopeRerank offers the GTE-Rerank series of text rerank models developed by Alibaba Tongyi Lab. You can integrate these models using the LlamaIndex framework to perform high-quality text retrieval and sorting.

Get started

Prerequisites

Activate Alibaba Cloud Model Studio. For more information, see Product Activation.
Create an API key. For more information, see Get and configure an API key.
Install the DashScopeRerank package (Python >=3.9 and <=3.12).

pip install llama-index-core
pip install llama-index-postprocessor-dashscope-rerank

Sample code

Set the API key.

export DASHSCOPE_API_KEY=YOUR_DASHSCOPE_API_KEY

from llama_index.core.data_structs import Node
from llama_index.core.schema import NodeWithScore
from llama_index.postprocessor.dashscope_rerank import DashScopeRerank

nodes = [
    NodeWithScore(node=Node(text="text1"), score=0.7),
    NodeWithScore(node=Node(text="text2"), score=0.8),
]

dashscope_rerank = DashScopeRerank(top_n=5)
results = dashscope_rerank.postprocess_nodes(nodes, query_str="<user query>")
for res in results:
    print("Text: ", res.node.get_content(), "Score: ", res.score)

Sample code output

Text:  text1 Score:  0.25589250620997755
Text:  text2 Score:  0.18071043165292258

Parameters

Input parameters

The following are the parameters for DashScopeRerank.

Parameter	Type	Default value	Description
model	string	gte-rerank	Supported models: gte-rerank
top_n	int	3	The number of top-ranked documents to return. If this parameter is not specified, or if its value is greater than the number of candidate documents, all documents are returned.
return_documents	bool	False	Specifies whether to include the original document text in the sorted results. The default value is `False`.
api_key	str	None	The DashScope API key. This can be set through an environment variable or other methods.