2024 GitHub Accelerator: Meet the 11 projects shaping open source AI

I was excited for Accelerator. All of the projects from last year[1] actually looked useful or I was already using them (htmx, dioxus, nuxt, datasette, trpc, strawberryql). Now it is a bunch of garbage LLM tools I will never use. Even their front page[2] looks infinitely worse than I remember and only says &quot;Powering AI Advancements in the open.&quot; RIP GitHub Accelerator.<p>EDIT: they really did make the front page much worse[3]. The words on the original page actually sound like they were written by a human. No mention of uplifting the open source ecosystem anymore, not even pictures of OSS maintainers.<p>[1]: <a href="https:&#x2F;&#x2F;github.blog&#x2F;2023-04-12-github-accelerator-our-first-cohort-and-whats-next&#x2F;" rel="nofollow">https:&#x2F;&#x2F;github.blog&#x2F;2023-04-12-github-accelerator-our-first-...</a> [2]: <a href="https:&#x2F;&#x2F;accelerator.github.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;accelerator.github.com&#x2F;</a> [3]: <a href="https:&#x2F;&#x2F;archive.is&#x2F;PIqUT" rel="nofollow">https:&#x2F;&#x2F;archive.is&#x2F;PIqUT</a>

It&#x27;s a privilege to be part of this years accelerator, especially given how awesome all the other projects are

This year&#x27;s projects are very exciting - here is a list of the projects:<p>Unsloth AI - Bending the cost curve of finetuning models Aframe (<a href="https:&#x2F;&#x2F;aframe.io&#x2F;" rel="nofollow">https:&#x2F;&#x2F;aframe.io&#x2F;</a>) - Making AR&#x2F;VR more performant in the broswer Giskard (<a href="https:&#x2F;&#x2F;lnkd.in&#x2F;gpmnjxs8">https:&#x2F;&#x2F;lnkd.in&#x2F;gpmnjxs8</a>) - Enterprise Testing platform for AI models Langdrive (<a href="https:&#x2F;&#x2F;addy.so&#x2F;" rel="nofollow">https:&#x2F;&#x2F;addy.so&#x2F;</a> and <a href="https:&#x2F;&#x2F;docs.langdrive.ai&#x2F;" rel="nofollow">https:&#x2F;&#x2F;docs.langdrive.ai&#x2F;</a>) - Easy to train LLM with API - Plug and play APIs for LLMs LLMWare (by Ai Bloks) (<a href="https:&#x2F;&#x2F;lnkd.in&#x2F;geS_nJu3">https:&#x2F;&#x2F;lnkd.in&#x2F;geS_nJu3</a>) - Simplifying the way enterprises use Small Language Models Open WebUI (<a href="https:&#x2F;&#x2F;openwebui.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;openwebui.com&#x2F;</a>) - Making LLMs available to anyone locally via UI to bring privacy, security and performance talkd.ai (<a href="https:&#x2F;&#x2F;lnkd.in&#x2F;gg5GPQ">https:&#x2F;&#x2F;lnkd.in&#x2F;gg5GPQ</a>) - Optimizing LLMs with Easy RAG deployment and management Web-Check (<a href="https:&#x2F;&#x2F;web-check.xyz&#x2F;" rel="nofollow">https:&#x2F;&#x2F;web-check.xyz&#x2F;</a>) - AI powered security insights bawed on open data from any website HackingbuddyGPT (<a href="https:&#x2F;&#x2F;hackingbuddy.ai&#x2F;" rel="nofollow">https:&#x2F;&#x2F;hackingbuddy.ai&#x2F;</a>) - copilots for security teams with autonomous agents marimo (<a href="https:&#x2F;&#x2F;marimo.io&#x2F;" rel="nofollow">https:&#x2F;&#x2F;marimo.io&#x2F;</a>) - raising the bar for ML and data science notebooks Nav2 (<a href="https:&#x2F;&#x2F;nav2.org&#x2F;" rel="nofollow">https:&#x2F;&#x2F;nav2.org&#x2F;</a>) - home of robotics navigation

Is it just me or are all the RAG workflow tools doing 80% of the same thing? I don&#x27;t know why I need them for a workflow.<p>I see only one true AI tool in the list: unsloth&#x2F;unsloth