OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

Deepke

DeepKE is an open-source Python toolkit for extracting entities, relations and attributes from text and multimodal data to support knowledge graph construction. Its documented scenarios include cnSchema, low-resource and document-level extraction, with examples for supervised training and prediction.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Deepke — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use DeepKE when a knowledge graph workflow needs structured entity, relation or attribute extraction and the team can work with its Python-based examples, data preparation and model setup.

Who it is for

  • Developers building knowledge graph ingestion workflows from text or multimodal inputs.
  • Researchers and practitioners training or applying named-entity, relation, attribute or event extraction models.

Documented uses

  • Identify and classify named entities in unstructured text.
  • Extract relationships between entities and attributes associated with them.
  • Train a supervised relation extraction model on prepared data and run prediction with a trained checkpoint.
  • Explore documented extraction approaches for cnSchema, low-resource, document-level and multimodal scenarios.
Before you choose

Requirements & limits

The project is a public GitHub repository with an MIT license. Its README links to documentation, an online demo and a Colab notebook.

  • The standard manual setup specifies Python 3.8 and lists task dependencies; DeepKE-LLM uses a separate environment.
  • The README recommends Linux and gives a Windows file-path note.
  • Training requires task data, which may be downloaded from an example or prepared using the project's annotation guidance.
  • The repository documents Docker environment setup as an alternative to manual configuration.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run prediction with a trained relation extraction model

Use the documented relation extraction example to run prediction after training.

Before you start

  • The DeepKE environment and relation extraction example described in its repository
  • A trained model checkpoint and prediction configuration

Steps

  1. Open the prediction configuration at DeepKE/example/re/standard/predict.yaml.

  2. Set the trained model path in the configuration, using the absolute path as directed by the README.

  3. Run python predict.py from the relation extraction example directory.

Expected outcome

The example's prediction script runs with the selected trained checkpoint and configuration.

Guide limitations

  • The guide assumes a trained model and does not cover how to evaluate or deploy prediction results.
  • The README recommends Linux; Windows users are advised to use backslashes in file paths.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
14 October 2026
Original website title
zjunlp/DeepKE · Official homepage title · retained in its source language
How we review sourcesSuggest a correction