DashScopeCloudRetriever

Get started

Prerequisites

Log on to https://bailian.console.aliyun.com/cn-beijing/ and obtain your API key. To specify a workspace, obtain the workspace ID.

An API key is created. For more information, see Get and configure an API key.
A workspace is created. For more information, see Workspace management.

Then, install the DashScopeCloudIndex package, which requires Python versions 3.9 to 3.12.

pip install llama-index-core
pip install llama-index-indices-managed-dashscope

import os
from llama_index.indices.managed.dashscope.retriever import DashScopeCloudRetriever

os.environ["DASHSCOPE_API_KEY"] = "your_api_key_here"
os.environ["DASHSCOPE_WORKSPACE_ID"] = "your_workspace_here"

retriever = DashScopeCloudRetriever("your index name")
nodes = retriever.retrieve("test query")
print(nodes)

Parameter descriptions

Input parameters

Parameter	Type	Default value	Description
index_name	str	None	The index name.
dense_similarity_top_k	int	100	The number of items to retrieve from vector search.
sparse_similarity_top_k	int	100	The number of items to retrieve from text search.
enable_reranking	bool	True	Specifies whether to use a reranking model.
rerank_model_name	str	gte-rerank-hybrid	The name of the reranking model. Supported models include the following: gte-rerank-hybrid gte-rerank
rerank_min_score	float	0.0	The score threshold for filtering retriever nodes. The retriever returns only nodes with a score higher than rerank_min_score. This parameter takes effect only when enable_reranking is set to True.
rerank_top_n	int	5	The number of nodes that the retriever returns. If the specified value exceeds the number of candidate nodes, all candidate nodes are returned.
workspace_id	str	None	The ID of the DashScope workspace. You can set this using an environment variable or other methods.
api_key	str	None	The DashScope API key. You can set this using an environment variable or other methods.

​Get started

​Prerequisites

​Parameter descriptions

​Input parameters

Get started

Prerequisites

Parameter descriptions

Input parameters