There was one time I was tracking down a race condition in a multithreaded code, and I forced it to run in a single thread so I have easier time to understand it.
This sped the code up by 30%, AND it fixed the bug.
So yeah, multithreading is hard!
@Notsogodboi@bee_fumo Speaking from experience from something like 6 years ago when I didn't know much about linux, it was marketed as a beginner friendly version of arch
@RyDawgE_@Jonathan_Blow@videolan Why don't people use mpv? I've had less issues with it than vlc (which i barely had issues with in the first place) and its a lot snappier
@chaotictransfem Would it though? Don't these models already use randomness in their outputs so they already select "suboptimal" tokens anyway. How would this be any different?
@jason_haugh@Pv@max_paperclips How is this tone deaf, virtually every ai company will do the same
"Examples for Section 1 include: Aleph Alpha, Anthropic, Black Forest Labs, Cohere, Google, Meta, Microsoft, Mistral, Open AI, Synthesia"
https://t.co/LTAetmagKn
@vaxryy I've been using zed with ai turned off for a year or so and it's pretty neat. Basically vsc but slightly faster and less ram hungry. Fewer extensions might be a deal breaker depending on what you're used to tho ig