525: Beating Apple to the Sauce

We daily drive Asahi Linux on a MacBook, chat about how the team beat Apple to a major GPU milestone, and an easy way to self-host open-source ChatGPT alternatives.Special Guest: Neal Gompa.Sponsored By:Tailscale: Tailscale is a programmable networking software that is private and secure by default - get it free on up to 100 devices!Linode Cloud Hosting: A special offer for all Linux Unplugged Podcast listeners and new Linode customers, visit linode.com/unplugged, and receive $100 towards your new account. Kolide: Kolide is a device trust solution for companies with Okta, and they ensure that if a device isn’t trusted and secure, it can’t log into your cloud apps.Support LINUX UnpluggedLinks:🎉 Alby — Boost into the show, first grab Alby, top it off, and then head over to the Podcast Index.⚡️ LINUX Unplugged on the Podcastindex.org — You can boost from the web. Once Alby is topped off, visit our page on the Podcast Index.Hector Martin's Controversial Question — Would you be okay with us adding some really trivial telemetry to the Asahi installer?Berlin with Brent — Brent will be back in Berlin for the Nextcloud Conference and can't get enough of Berlin Meetups! Friday, September 8th, 6 PM.Fedora Asahi RemixFedora Asahi Remix Coming For Fedora Linux On Apple Silicon Hardware — Fedora Asahi Remix will be their new flagship distribution for providing a polished Linux experience on Apple Silicon.Fedora Asahi Remix: bringing Fedora to Apple Silicon Macs (Flock To Fedora 2023)Our new flagship distro: Fedora Asahi Remix — We’re still working out the kinks and making things even better, so we are not quite ready to call this a release yet. We aim to officially release the Fedora Asahi Remix by the end of August 2023. Look forward to many new features, machine support, and more!Hector Martin: “Okay, I’m going to be honest…” — I apologize to all Asahi Linux users. You deserve better. When I chose Arch Linux ARM as a base I didn't realize it would have so many basic QA issues.Coming soon: Fedora for Apple Silicon Macs! (Fedora Discourse)The first conformant M1 GPU driver — Our reverse-engineered, free and open source graphics drivers are the world’s only conformant OpenGL ES 3.1 implementation for M1- and M2-family graphics hardware. That means our driver passed tens of thousands of tests to demonstrate correctness and is now recognized by the industry.Asahi Linux’s Apple M1/M2 Gallium3D Driver Now OpenGL ES 3.1 Conformant — It's even more rewarding for the community developers in that Apple doesn't provide any conformant (OpenGL or Vulkan) graphics drivers for their Arm-based platform.Feature Support · AsahiLinux/docs WikiSwitch to the kernel-16k variant - Fedora DiscussionNixOS: Unlocking your LUKS via SSH and TorStreetComplete — Easy to use OpenStreetMap editor for Android.getumbrel/llama-gpt: A self-hosted, offline, ChatGPT-like chatbot. Powered by Llama 2. 100% private, with no data leaving your device.serge-chat/serge: A web interface for chatting with Alpaca through llama.cpp. Fully dockerized, with an easy to use API.liltom-eth/llama2-webui: Run any Llama 2 locally with gradio UI on GPU or CPU from anywhere (Linux/Windows/Mac). Usellama.cpp — Port of Facebook’s LLaMA model in C/C++Llama2.c — Inference Llama 2 in one file of pure CKoboldcpp — A simple one-file way to run various GGML models with KoboldAI’s UIlollms-webui — Lord of Large Language Models Web User InterfaceLM Studio — Discover, download, and run local LLMstext-generation-webui — A Gradio web UI for Large Language Models. Supports transformers, GPTQ, llama.cpp (ggml/gguf), Llama models.A comprehensive guide to running Llama 2 locally — Code Llama is a state-of-the-art LLM capable of generating code, and natural language about code, from both code and natural language prompts.Meta Releases Code Llama, a Coding Version of Llama 2Introducing Code Llama, a state-of-the-art large language model for codingLlama and ChatGPT Are Not Open-SourceMeta launches Llama 2, a source-available AI model that allows commercial applications — A family of pretrained and fine-tuned language models in sizes from 7 to 70 billion parameters.Meta’s Llama 2 is not open source — Meta's newly released large language model Llama 2 is not open source.