Scalars, vectors, matrices... Tensors are just numbers organized by shape. They're the numerical representation of your input that every ML model operates on.
๐ค Transformers.js lets you run state-of-the-art machine learning directly from JavaScript
ONNX, onnxruntime, model files, caching, tensors, and pipelines in 30 seconds โฐ
This setup is basically my daily driver, a set of tools, chainable, fully interoperable and thus allowing you to create entire complex pipelines out of one prompt. From ideations, implementation to research, review and verification โค๏ธ
New from KERN: say hi to Agon.
A CLI that runs your coding task across every AI engine you have at once, in isolated git worktrees, scores them, and applies the best test-passing patch. Glicko-2 ratings route the next task to whoever actually earns it.
Any AI can join. They compete. You ship.
And next: agonloop. Turning Agon into a real, continuous loop used by other CLI like @claudeai or @OpenAI's Codex. Coming soon.
https://t.co/UpBts3ocbg
https://t.co/0DDKdDQEKY
New from KERN: say hi to Agon.
A CLI that runs your coding task across every AI engine you have at once, in isolated git worktrees, scores them, and applies the best test-passing patch. Glicko-2 ratings route the next task to whoever actually earns it.
Any AI can join. They compete. You ship.
And next: agonloop. Turning Agon into a real, continuous loop used by other CLI like @claudeai or @OpenAI's Codex. Coming soon.
https://t.co/UpBts3ocbg
https://t.co/0DDKdDQEKY
@anshnanda i tried it by making a small websocket server where each connecting agent and me, the human all could chat, sometimes claude sessions really were invested but often they just read the messages, replied and then waited for input again :(
downloading a store/launcher update, to download a game update, to then download a ingame asset update, so i can play a cardgame ... this is completely normal UX and exactly what i am used to when i want to play a online game. thanks @MTG_Arena