Shatur@discuss.tchncs.deEnglish · 10 hours agoNew NVIDIA paper says AI models do not always need long contextplus-squarearxiv.orgexternal-linkmessage-square5linkfedilinkarrow-up126arrow-down13
arrow-up123arrow-down1external-linkNew NVIDIA paper says AI models do not always need long contextplus-squarearxiv.orgShatur@discuss.tchncs.deEnglish · 10 hours agomessage-square5linkfedilink
absGeekNZ@lemmy.nzEnglish · 1 day agoOS recommendationsplus-squaremessage-squaremessage-square20linkfedilinkarrow-up116arrow-down15
arrow-up111arrow-down1message-squareOS recommendationsplus-squareabsGeekNZ@lemmy.nzEnglish · 1 day agomessage-square20linkfedilink
Spite_By_Design_LLC@lemmy.dbzer0.comEnglish · 1 day agoBritish fintech Revolut slashes AI coding costs 10x by cutting out Claude and going self-hostedplus-squarewww.theregister.comexternal-linkmessage-square17linkfedilinkarrow-up169arrow-down13
arrow-up166arrow-down1external-linkBritish fintech Revolut slashes AI coding costs 10x by cutting out Claude and going self-hostedplus-squarewww.theregister.comSpite_By_Design_LLC@lemmy.dbzer0.comEnglish · 1 day agomessage-square17linkfedilink
Shimitar@downonthestreet.euEnglish · 11 hours agoVideo cards, multi GPUs questionplus-squaremessage-squaremessage-square4linkfedilinkarrow-up16arrow-down11
arrow-up15arrow-down1message-squareVideo cards, multi GPUs questionplus-squareShimitar@downonthestreet.euEnglish · 11 hours agomessage-square4linkfedilink
Shatur@discuss.tchncs.deEnglish · 2 days agoChinese AI Writes Weaker Code for Beijing’s Adversaries. We Fixed It. A Qwen3-Coder Case Studyplus-squarewww.hirundo.ioexternal-linkmessage-square5linkfedilinkarrow-up143arrow-down16
arrow-up137arrow-down1external-linkChinese AI Writes Weaker Code for Beijing’s Adversaries. We Fixed It. A Qwen3-Coder Case Studyplus-squarewww.hirundo.ioShatur@discuss.tchncs.deEnglish · 2 days agomessage-square5linkfedilink
Nzo@lemmy.zipEnglish · 2 days agoEight LLMs (four open-weight) live in a 24/7 reality show; small open models moderate every lineplus-squarelemmy.zipexternal-linkmessage-square13linkfedilinkarrow-up139arrow-down112
arrow-up127arrow-down1external-linkEight LLMs (four open-weight) live in a 24/7 reality show; small open models moderate every lineplus-squarelemmy.zipNzo@lemmy.zipEnglish · 2 days agomessage-square13linkfedilink
Shimitar@downonthestreet.euEnglish · 2 days agoThoughts on coding with AI: my experienceplus-squarewiki.gardiol.orgexternal-linkmessage-square5linkfedilinkarrow-up128arrow-down15
arrow-up123arrow-down1external-linkThoughts on coding with AI: my experienceplus-squarewiki.gardiol.orgShimitar@downonthestreet.euEnglish · 2 days agomessage-square5linkfedilink
ThelastfingerofH@lemmy.worldEnglish · 3 days agoMythos-class model on huggingface: 14T parametersplus-squarehuggingface.coexternal-linkmessage-square12linkfedilinkarrow-up128arrow-down19
arrow-up119arrow-down1external-linkMythos-class model on huggingface: 14T parametersplus-squarehuggingface.coThelastfingerofH@lemmy.worldEnglish · 3 days agomessage-square12linkfedilink
BeefAndPoultry@lemmus.orgEnglish · 5 days agoHeretic 2.0 released: Model modification for everyone - "it can suppress slop, reduce sycophancy, alter emotional alignment, and pretty much anything else that you can think of"plus-squaregithub.comexternal-linkmessage-square6linkfedilinkarrow-up155arrow-down14
arrow-up151arrow-down1external-linkHeretic 2.0 released: Model modification for everyone - "it can suppress slop, reduce sycophancy, alter emotional alignment, and pretty much anything else that you can think of"plus-squaregithub.comBeefAndPoultry@lemmus.orgEnglish · 5 days agomessage-square6linkfedilink
BeatTakeshi@lemmy.worldEnglish · edit-24 days agoI have a RTX A1000 (8GB) with 64GB DRR5, what would you recommend for a newbie?plus-squaremessage-squaremessage-square32linkfedilinkarrow-up125arrow-down12
arrow-up123arrow-down1message-squareI have a RTX A1000 (8GB) with 64GB DRR5, what would you recommend for a newbie?plus-squareBeatTakeshi@lemmy.worldEnglish · edit-24 days agomessage-square32linkfedilink
Shatur@discuss.tchncs.deEnglish · 7 days agoSam Altman: "Local inference is a terrible allocation of scarce resources"plus-squaretwitee.coexternal-linkmessage-square175linkfedilinkarrow-up1588arrow-down112
arrow-up1576arrow-down1external-linkSam Altman: "Local inference is a terrible allocation of scarce resources"plus-squaretwitee.coShatur@discuss.tchncs.deEnglish · 7 days agomessage-square175linkfedilink
BeefAndPoultry@lemmus.orgEnglish · edit-25 days agoApparently Gemma 4's chat template can take over 10 seconds just to build the text when there are many tool callsplus-squarehuggingface.coexternal-linkmessage-square2linkfedilinkarrow-up122arrow-down14
arrow-up118arrow-down1external-linkApparently Gemma 4's chat template can take over 10 seconds just to build the text when there are many tool callsplus-squarehuggingface.coBeefAndPoultry@lemmus.orgEnglish · edit-25 days agomessage-square2linkfedilink
Spite_By_Design_LLC@lemmy.dbzer0.comEnglish · edit-24 days agoStarting from Dec, Codex and Chat will share same token pool. Thanks Sam! EABODplus-squaremessage-squaremessage-square15linkfedilinkarrow-up114arrow-down12
arrow-up112arrow-down1message-squareStarting from Dec, Codex and Chat will share same token pool. Thanks Sam! EABODplus-squareSpite_By_Design_LLC@lemmy.dbzer0.comEnglish · edit-24 days agomessage-square15linkfedilink
Shatur@discuss.tchncs.deEnglish · 7 days agoAnthropic quietly bans recreating existing games in Claude Codeplus-squaretwitee.coexternal-linkmessage-square35linkfedilinkarrow-up1125arrow-down18
arrow-up1117arrow-down1external-linkAnthropic quietly bans recreating existing games in Claude Codeplus-squaretwitee.coShatur@discuss.tchncs.deEnglish · 7 days agomessage-square35linkfedilink
Shatur@discuss.tchncs.deEnglish · 6 days agoFriends Don't Let Friends Use Ollamaplus-squaresleepingrobots.comexternal-linkmessage-square15linkfedilinkarrow-up1107arrow-down15
arrow-up1102arrow-down1external-linkFriends Don't Let Friends Use Ollamaplus-squaresleepingrobots.comShatur@discuss.tchncs.deEnglish · 6 days agomessage-square15linkfedilink
BeefAndPoultry@lemmus.orgEnglish · 8 days agoMistral Says Its AI Model ‘Le Chonk’ Has Leapfrogged China in the Open-Weight Race (1T-A52B)plus-squarewww.wired.comexternal-linkmessage-square23linkfedilinkarrow-up144arrow-down14
arrow-up140arrow-down1external-linkMistral Says Its AI Model ‘Le Chonk’ Has Leapfrogged China in the Open-Weight Race (1T-A52B)plus-squarewww.wired.comBeefAndPoultry@lemmus.orgEnglish · 8 days agomessage-square23linkfedilink
Shatur@discuss.tchncs.deEnglish · 6 days agoQwen27b vs GPT 6plus-squarediscuss.tchncs.deimagemessage-square10linkfedilinkarrow-up127arrow-down17
arrow-up120arrow-down1imageQwen27b vs GPT 6plus-squarediscuss.tchncs.deShatur@discuss.tchncs.deEnglish · 6 days agomessage-square10linkfedilink
xoron@programming.devEnglish · 3 days agoBrowser-based WebLLM Chatbotplus-squaremessage-squaremessage-square0linkfedilinkarrow-up11arrow-down12
arrow-up1-1arrow-down1message-squareBrowser-based WebLLM Chatbotplus-squarexoron@programming.devEnglish · 3 days agomessage-square0linkfedilink
Shatur@discuss.tchncs.deEnglish · 7 days agoSomeone built a 32-bit RISC-V CPU directly inside Qwen's model weights so it can execute instructions internaly!plus-squarediscuss.tchncs.devideomessage-square3linkfedilinkarrow-up138arrow-down16
arrow-up132arrow-down1videoSomeone built a 32-bit RISC-V CPU directly inside Qwen's model weights so it can execute instructions internaly!plus-squarediscuss.tchncs.deShatur@discuss.tchncs.deEnglish · 7 days agomessage-square3linkfedilink
sriram1983007@sh.itjust.worksEnglish · 5 days agoFine-tuned Qwen3-4B locally for one narrow decision — the stablecoin risk-gate side works, the dispute side is where it falls apartplus-squaresh.itjust.worksimagemessage-square1linkfedilinkarrow-up18arrow-down15
arrow-up13arrow-down1imageFine-tuned Qwen3-4B locally for one narrow decision — the stablecoin risk-gate side works, the dispute side is where it falls apartplus-squaresh.itjust.workssriram1983007@sh.itjust.worksEnglish · 5 days agomessage-square1linkfedilink