misk@sopuli.xyz to Technology@lemmy.worldEnglish · 10 months agoByteDance is secretly using OpenAI’s tech to build a competitorwww.theverge.comexternal-linkmessage-square5fedilinkarrow-up178arrow-down16cross-posted to: world@lemmy.world
arrow-up172arrow-down1external-linkByteDance is secretly using OpenAI’s tech to build a competitorwww.theverge.commisk@sopuli.xyz to Technology@lemmy.worldEnglish · 10 months agomessage-square5fedilinkcross-posted to: world@lemmy.world
minus-squarecyd@lemmy.worldlinkfedilinkEnglisharrow-up14arrow-down1·10 months agoStrange that they don’t just use an open weights model; there are several now that surpass ChatGPT 3.5, which is probably good enough for what they need.
minus-squareFaceDeer@kbin.sociallinkfedilinkarrow-up12·10 months agoMight be that they started training before those open models were available. Or they were just lazy and OpenAI’s API was easier.
minus-squaremuntedcrocodile@lemmy.worldlinkfedilinkEnglisharrow-up4·10 months agoWhat models have u foubd to surpass 3.5? Any outpaced gpt4 yet?
minus-squarecyd@lemmy.worldlinkfedilinkEnglisharrow-up5arrow-down1·10 months agoMistral 7B and deepseek-ai are two open-weight models that surpass 3.5, though not 4, on several measures.
Strange that they don’t just use an open weights model; there are several now that surpass ChatGPT 3.5, which is probably good enough for what they need.
Might be that they started training before those open models were available. Or they were just lazy and OpenAI’s API was easier.
What models have u foubd to surpass 3.5? Any outpaced gpt4 yet?
Mistral 7B and deepseek-ai are two open-weight models that surpass 3.5, though not 4, on several measures.