@jrysana Im really curious, how do you create prompts for such uncontaminated benchmark? Do you come up with the tasks on your own, or do you use any help like LLMs, papers, or 3rd party?
@Lonely__MH@opencode Bugs happen in all software. Look at the overall, opencode provides one of the best caching ratios, open to all providers, open source, community plugins, free inference. what else do you want?
@3M4C5@superdoteng Easier and quicker navigation, easier for eyes, native handling of text navigation and shortcuts, quicker for working and switching between multiple projects or multiple agents working at the same time.
Those are my most important reasons
@0xSero Oh shit you live in Warsaw? I live 40 min car drive from Warsaw. Im not interested in the rig, just surprised.
What made you choose to live in Poland? Also do you speak any polish? its a hard language
@0xSero I hate the term AGI, since public misunderstands it and there are so many diff definitions. AI is also a dumb term since its way too broad. Even an NPC is an AI even tho it doesn't use a neural network model
I wish society used the term LM or LLM, but it doesn't sell as well
@stamparm I made a similar project, it scrapes from many many sources and spits out only working proxies.
Maybe you find it useful, it's at https://t.co/7hkMDHLE0g
Such a powerful editor running entirely in a browser. It even supports all the extensions and syncs them with the VSCode app thanks to the GitHub account login. #HappyCoding via @code