Code Llama vs. StarCoder vs. Wing Python IDE vs. blanket.js Comparison Chart

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.

A seamless JavaScript code coverage library. Blanket.js is a code coverage tool for JavaScript that aims to be easy to install, easy to use, and easy to understand. Blanket.js can be run seamlessly or can be customized for your needs. JavaScript code coverage compliments your existing JavaScript tests by adding code coverage statistics (which lines of your source code are covered by your tests). Parsing the code using Esprima and node-falafel, and instrumenting the file by adding code tracking lines. Connecting to hooks in the test runner to output the coverage details after the tests have been completed. A Grunt plugin has been created to allow you to use Blanket like a "traditional" code coverage tool (creating instrumented copies of physical files, as opposed to live-instrumenting). Runs the QUnit-based Blanket report headlessly using PhantomJS. Results are displayed on the console, and the task will cause Grunt to fail if any of your configured coverage thresholds are not met.