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.

Self-hosted Whisper-based voice recognition server for open Android phones

15 pointsby nichohelover 2 years ago

2 comments

smoldesuover 2 years ago
I suspect something similar is possible with ChatGPT. Using the GPT-neo-125m model I&#x27;ve been able to get some really convincing (if lackluster) answers on 4 core ARM hardware and less than 2gb of memory. With enough sampling, you can get legible paragraph-length responses out in less than 10 seconds; that&#x27;s pretty good for an offline program in my book.<p>I&#x27;m using rust-bert to serve it over a Discord bot, similar to one of their examples[0]. It&#x27;s running on Oracle VCPUs right now, but with dedi hardware and ML acceleration I bet it would scream!<p>[0] <a href="https:&#x2F;&#x2F;github.com&#x2F;guillaume-be&#x2F;rust-bert&#x2F;blob&#x2F;master&#x2F;examples&#x2F;generation_gpt_neo.rs">https:&#x2F;&#x2F;github.com&#x2F;guillaume-be&#x2F;rust-bert&#x2F;blob&#x2F;master&#x2F;exampl...</a>
评论 #34784951 未加载
nichohelover 2 years ago
With this little bit of code you can use excellent voice recognition (ggerganov whisper.cpp port of Whisper) hosted on your own server, for your de-Googled Android phone, for text messaging, emails, search, and so on.