inari@piefed.zip to Technology@lemmy.worldEnglish · edit-25 months agoDeepSeek ditches Nvidia for Huawei chips in V4 launchcybernews.comexternal-linkmessage-square86linkfedilinkarrow-up1410arrow-down18
arrow-up1402arrow-down1external-linkDeepSeek ditches Nvidia for Huawei chips in V4 launchcybernews.cominari@piefed.zip to Technology@lemmy.worldEnglish · edit-25 months agomessage-square86linkfedilink
minus-squarebrucethemoose@lemmy.worldlinkfedilinkEnglisharrow-up4arrow-down1·5 months agoI just meant for mass inference serving. Yeah, I haven’t seen much in the way of bitnet training savings yet, like regular old QAT. It does appear that Deepseek is finetuning their MoEs in a 4-bit format now, though.
I just meant for mass inference serving.
Yeah, I haven’t seen much in the way of bitnet training savings yet, like regular old QAT. It does appear that Deepseek is finetuning their MoEs in a 4-bit format now, though.