Recently, the CodeFuse IDE 0.6 version was officially released, introducing editor diagnostics and AI fix features, as well as supporting the simultaneous appearance of inline completions during dropdown suggestions, further enhancing the convenience and efficiency of code writing. CodeFuse IDE is an AI integrated development environment (IDE) developed based on Ant's self-developed large model CodeFuse and its own IDE framework OpenSumi, supporting mainstream programming languages and providing code assistance during the development process.

["Ant Group CodeFuse announces the open-source DevOps-ChatBot end-to-end AI intelligent assistant","DevOps-ChatBot is an open-source end-to-end intelligent AI assistant designed specifically for the entire lifecycle of software development","The project supports private deployment to ensure data security and high availability for specific tasks","By integrating technologies such as knowledge bases and knowledge graph enhancements, it ensures the accuracy and timeliness of generated content","The project also supports interactive code modification, compilation, and execution"]

Ant Group's CodeFuse Plugin for Development Assistant has released a new version, adding support for Android Studio. The CodeFuse plugin is now compatible with 11 mainstream IDEs, including Visual Studio Code and IDEA. The new version optimizes the code completion accuracy for front-end languages such as JavaScript and TypeScript. New features include a 'Dev Assistant' product Q&A entry and a logout function. Developers can apply for testing qualifications on the official website.

The ModelCache architecture under Ant Group's CodeFuse includes the adapter, embedding, similarity, and data_manager modules. ModelCache can reduce the inference cost of large model applications and enhance user experience. Cache hits can reduce average latency by 10 times, with speed improvements of up to 14.5%. ModelCache will continue to optimize performance and accuracy, improving recall time.

Ant Group has released the 4bits quantized version of CodeFuse-CodeLlama-34B. The quantized model can be loaded on a single A10 or RTX4090. The quantized model achieves a performance of 73.8% on the Humaneval pass@1 metric. CodeFuse is Ant's self-developed large model for code generation, aimed at enhancing developers' efficiency by providing intelligent suggestions and real-time support.