March 27, 2012 - In 2025 HuaweiDuring the China Partner Conference, theHuawei ModelEngine AI Full Process Toolchain to Global DevelopersOpen Sourcerelease.
The open source version of ModelEngine builds a full process AI tool chain around data enablement, model enablement, and application enablement.Committed to solving the key problems of time-consuming data engineering, difficult model training and application landing encountered in AI industrialization.The company's AI application is designed to shorten the overall cycle of training a general-purpose big model into an industry big model based on industry private domain data and developing it into an AI application, accelerating the landing of AI in the industry.
ModelEngine open source version of the built-in data processing algorithms , a complete model management process , application-oriented to provide low-code orchestration , RAG framework , custom plug-in capabilities .
- Data Enable:Toolchain for summarizing and processing various types of data such as text, images, documents, etc. It provides key capabilities such as data cleansing, data evaluation, QA pair generation, knowledge generation, etc., and provides corpus and knowledge for large model training and RAG applications.
- Model Enable:A toolchain for model deployment, training, fine-tuning and inference, with one-click operation to improve model training and push efficiency.
- Application Enable:Provides a one-stop development, debugging and publishing tool chain for AI applications, helping to efficiently build AI applications and quickly validate their effectiveness.
ModelEngine open source code is hosted on multiple platforms, 1AI with open source links:
- gitcode:https://gitcode.com/ModelEngine
- gitee:https://gitee.com/openeuler/modelengine
- github:https://github.com/ModelEngine-Group
