English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Automatically generating source code from natural language descriptions hasbeen a growing field of research in recent years. However, current large-scalecode generation models often encounter difficulties when selecting appropriateAPIs for specific contexts. These models may generate APIs that do not meetrequirements or refer to non-existent APIs in third-party libraries, especiallyfor lesser-known or private libraries. Inspired by the process of humandevelopers using tools to search APIs, we propose ToolCoder, a novel approachthat integrates API search tools with existing models to assist in codegeneration and API selection. To teach our model to use tools, we introduce anautomated data annotation method using ChatGPT to add tool usage informationinto the source code data and fine-tune code generation models. Duringinference, we integrate API search tools into the generation process so thatour model can automatically use the search tool to get suggestions whenselecting an API. Our experimental results demonstrate that ToolCoder exhibitsexcellent performance and generalization across five public and private librarycode generation benchmarks, with at least 6.21\% improvement on average pass@1metrics and 9.64\% improvement on average pass@10 metrics compared tostate-of-the-art methods. Furthermore, we show that our relatively smallToolCoder model is comparable to one of the current best models, GPT-3.5,highlighting the potential of incorporating programming tools into the codegeneration process.

ToolCoder: Teach Code Generation Models to use API search tools