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: Tabby – AI Coding Assistant Runs on Apple M1/M2 GPU

3 pointsby wsxiaoysover 1 year ago
Hi HN, after five months of hard work since Tabby&#x27;s previous Show HN (<a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=35470915">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=35470915</a>), I&#x27;m thrilled to share you the release of Tabby v0.1.1.<p>Previously, Tabby ran exclusively on CUDA devices, posing a significant barrier for developers looking to effectively utilize LLMs in their day-to-day coding.<p>The Tabby team made a significant contribution by enhancing support for the StarCoder series models (1B&#x2F;3B&#x2F;7B) in llama.cpp. This enhancement allows these models to run on Metal, providing comparable performance to that of an NVIDIA GPU.<p>I eagerly look forward to receiving feedback from the community and witnessing the enhanced on-edge deployment experience!<p>References:<p>[1] Apple Installation Instructions: <a href="https:&#x2F;&#x2F;tabby.tabbyml.com&#x2F;docs&#x2F;installation&#x2F;apple&#x2F;" rel="nofollow noreferrer">https:&#x2F;&#x2F;tabby.tabbyml.com&#x2F;docs&#x2F;installation&#x2F;apple&#x2F;</a><p>[2] Launch blog: <a href="https:&#x2F;&#x2F;tabby.tabbyml.com&#x2F;blog&#x2F;2023&#x2F;09&#x2F;18&#x2F;release-0-1-1-metal&#x2F;" rel="nofollow noreferrer">https:&#x2F;&#x2F;tabby.tabbyml.com&#x2F;blog&#x2F;2023&#x2F;09&#x2F;18&#x2F;release-0-1-1-meta...</a>

2 comments

gyxlucyover 1 year ago
This is quite cool. Have you compared performance vs latency tradeoff between 1B&#x2F;3B&#x2F;7B models? What would you recommend as the ideal choice?
myinover 1 year ago
Cool to be able to run locally