#490 · Primary category： AI Coding Assistants
gpt-code-clippy
Full description can be found here: https://discuss.huggingface.co/t/pretrain-gpt-neo-for-open-source-github-copilot-model/7678?u=ncoop57
Project last updated：01/18/22
GitHub Stars
3.3K
Forks
217
Contributors
15
License
Apache-2.0
Why we included this project
GPT-Code-Clippy is one of the earliest public attempts to rebuild GitHub Copilot as an open, self-hostable project, and it is worth reading mainly as a teaching artifact. The repository shows how GPT-CC, a GPT-Neo and GPT-2 model fine-tuned on public GitHub code, was actually put together: the dataset scraping criteria, deduplication scripts, training hyperparameters, and HumanEval results are all documented in one place. Anyone planning to build a Copilot-style assistant will find the data preparation and fine-tuning pipeline worth studying, and the bundled Visual Studio Code extension demonstrates how the model plugs into an editor. Do not expect a maintained drop-in tool, though; treat this as a blueprint for understanding and reproducing this kind of code completion system on your own.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
superpowers
An agentic skills framework & software development methodology that works.
ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
skills
Skills for Real Engineers. Straight from my .agents directory.
opencode
The open source coding agent.
andrej-karpathy-skills
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.