Imaginary Programming vs. IntelliSense vs. Python vs. StarCoder Comparison Chart

IntelliSense is a general term for various code editing features including: code completion, parameter info, quick info, and member lists. IntelliSense features are sometimes called by other names such as "code completion", "content assist", and "code hinting." Visual Studio Code IntelliSense is provided for JavaScript, TypeScript, JSON, HTML, CSS, SCSS, and Less out of the box. VS Code supports word based completions for any programming language but can also be configured to have richer IntelliSense by installing a language extension. VS Code IntelliSense features are powered by a language service. A language service provides intelligent code completions based on language semantics and an analysis of your source code. If a language service knows possible completions, the IntelliSense suggestions will pop up as you type. If you continue typing characters, the list of members (variables, methods, etc.) is filtered to only include members containing your typed characters.

StarCoder and StarCoderBase are Large Language Models for Code (Code LLMs) trained on permissively licensed data from GitHub, including from 80+ programming languages, Git commits, GitHub issues, and Jupyter notebooks. Similar to LLaMA, we trained a ~15B parameter model for 1 trillion tokens. We fine-tuned StarCoderBase model for 35B Python tokens, resulting in a new model that we call StarCoder. We found that StarCoderBase outperforms existing open Code LLMs on popular programming benchmarks and matches or surpasses closed models such as code-cushman-001 from OpenAI (the original Codex model that powered early versions of GitHub Copilot). With a context length of over 8,000 tokens, the StarCoder models can process more input than any other open LLM, enabling a wide range of interesting applications. For example, by prompting the StarCoder models with a series of dialogues, we enabled them to act as a technical assistant.