TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Show HN: I reproduced Code Llama fill-in-the-middle code completion training

2 pointsby BohdanPetryshyn11 months ago
As part of my master&#x27;s thesis on optimizing LLMs for code completion in OpenAPI format, I implemented a script for fine-tuning Code Llama for Copilot-style code completion. Surprisingly, I was not able to find a similar implementation before since most works focus on instruction fine-tuning for chat-like interaction.<p>Related publication: <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2405.15729" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2405.15729</a> Code Llama fill-in-the-middle fine-tuning: <a href="https:&#x2F;&#x2F;github.com&#x2F;BohdanPetryshyn&#x2F;code-llama-fim-fine-tuning">https:&#x2F;&#x2F;github.com&#x2F;BohdanPetryshyn&#x2F;code-llama-fim-fine-tunin...</a> OpenAPI completion benchmark: <a href="https:&#x2F;&#x2F;github.com&#x2F;BohdanPetryshyn&#x2F;openapi-completion-benchmark">https:&#x2F;&#x2F;github.com&#x2F;BohdanPetryshyn&#x2F;openapi-completion-benchm...</a>

no comments

no comments