Skip to content

discussions Search Results · language:Dune language:Python language:JavaScript is:public language:JavaScript language:TypeScript

3.7M results  (526 ms)

3.7M results

The language qualifier is not supported when searching discussions. Try searching for repositories instead?

All GitHub Copilot plans are now on usage-based billing Hello GitHub Community, Usage-based billing is now in effect for all GitHub Copilot plans. Please refer to our initial announcement and today’s ...

I have this integration exposed to HomeKit, however when Arming or Disarming the Integration uses always the User Code 1 and wanted to set a Usercode Set in Panel Entity which should be is the User Code ...

I d like to suggest adding support for Google s Material Symbols, the successor to the standard Material Icons set. Beyond a broader icon selection, Material Symbols are variable fonts with configurable ...

Completions are generated using usage for mise and I expected the same to occur for fnox , but in my bash session completions do not work. My .bashrc has this at the end: eval $(mise activate bash) ...

The Copilot model picker with a pull-down menu showing MAI Code 1 Flash selected MAI‑Code‑1‑Flash, Microsoft s latest small‑tier coding model, is now rolling out in GitHub Copilot starting with VS Code. ...

I m an avid oMLX user, but my lab isn t oMLX-only — I also run LM Studio, llama.cpp, MLX (vanilla), and vLLM / Docker model runner on different nodes. I wanted one URL for coding agents across all of it, ...

Hyprland version 0.55.3-1 Regression yup Describe the bug pretty much as simple as the title says; the spiral.lua layout does not function properly on latest hyprland. errors are shown below img ...

Hey there, I m a product manager for IBM s Carbon Design System. I ve been using Open Design to prototype and I ve been really enjoying it. I noticed that you have Carbon listed in the design system plugins, ...

https://github.com/user-attachments/assets/603b1a7a-b957-4edd-8020-9c9cfecb0f7e For @aether1941

For Flux 2 generation under memory constrained hardware: Why load text encoder ~4GB and diffusor ~4-8GB at the same time? We don t need the TE loaded at all during generation. Just cache the embedding ...