21 lines
951 B
Markdown
21 lines
951 B
Markdown
|
---
|
||
|
|
type: post
|
||
|
|
draft: true
|
||
|
|
title: "Introducing Turbopilot"
|
||
|
|
date: 2023-04-16T14:08:55Z
|
||
|
|
description: My local intelligent auto-complete project
|
||
|
|
url: /2023/04/16/turbopilot
|
||
|
|
type: post
|
||
|
|
mp-syndicate-to:
|
||
|
|
- https://brid.gy/publish/mastodon
|
||
|
|
- https://brid.gy/publish/twitter
|
||
|
|
resources:
|
||
|
|
- name: feature
|
||
|
|
src: images/thesis_mug_small.jpg
|
||
|
|
tags:
|
||
|
|
- ai
|
||
|
|
- open-source
|
||
|
|
|
||
|
|
---
|
||
|
|
|
||
|
|
I started TurboPilot over the easter weekend when I was stuck at home bored with COVID. As an AI specialist, I've been following OpenAI, Copilot and all things GPT very closely and I've been enspired excited by all the open activity like [llama.cpp](https://github.com/ggerganov/llama.cpp) which allows you to run large language models locally on CPU. I decided that it might be quite useful to have intelligent code autocompletion that runs locally without sending my data to the OpenAI mothership for analysis and when I'm travelling with limited connectivity like on trains or planes.
|