swax (@swax) header
swax (@swax) profile picture

https://teapawt.net/t.co/tHM3YaNjAV

Joined February 2008
Photo by swax (@swax)Photo by swax (@swax)Photo by swax (@swax)Photo by swax (@swax)Photo by swax (@swax)Photo by swax (@swax)Photo by swax (@swax)Photo by swax (@swax)Photo by swax (@swax)Photo by swax (@swax)
A 1994 Japanese comedy, now in English, in the original cast's own voices. Each scene plays in Japanese, then in the AI dub. 680 lines and 21 characters, all dubbed on one home PC. 🎬
1
80
Garibaldi speed profiles
29
Testing SSX Tricky real-time lighting More vids here: youtube.com/playlist?list=PL…
72
Use your favorite AI harness to Dub Anything. Preserve the actors' own voices. All the audio processing is done on your local machine. Open source github.com/swax/dub-anything
1
83
Used Astra to port sketchtv.lol from Vercel to a Droplet. Migrated from Next.js to Rust/Astro. Used github.com/swax/jumpgate to generate this interactive code diagram to understand the architecture.
1
73
Creating a new SSX Tricky compatible map in VR/AR with OpenSlope github.com/swax/openslope
2
128
working on a real-time kernal/user space activity view follow up to code perspective
81
The Sheila Divine - Hum youtube.com/watch?v=wAg_zHed…
2
134
Be pleasantly surprised every few hours with a fresh wallpaper - generated randomly from your own ideas. Bring your own keys. github.com/swax/WallTrek
1
337
Latest version of UnityMCP-VRC allows multiple agents to work with Unity simultaneously. Also greatly improves the ability to wire up UdonSharp scripts. Plus new tools from @SetohimaVrc and paging support for long responses. github.com/swax/UnityMCP-VRC
136
naisys dashboard - managing agents dynamically distributed across machines/platforms/models
1
86
Very cool to post a large source file into ChatGPT and get back a visual diagram of how it all works. Makes familiarizing yourself with new code much easier.
1
88
ChatGPT Images 2.0 has its limits - Loaded in an Excel sheet of health care spending for it to visualize. It does a much better job than previous models, but doesn't generate a coherent diagram.
127