ylai@lemmy.ml to AI@lemmy.ml • 4 months agoNvidia’s ‘Nemotron-4 340B’ model redefines synthetic data generation, rivals GPT-4external-linkmessage-square2 fedilinkarrow-up110
arrow-up110external-linkNvidia’s ‘Nemotron-4 340B’ model redefines synthetic data generation, rivals GPT-4ylai@lemmy.ml to AI@lemmy.ml • 4 months agomessage-square2 Commentsfedilink
minus-squareFisch@discuss.tchncs.dehexbear1·4 months ago340B is fucking huge, holy shit. How big is GPT-4? linkfedilink
minus-squareylai@lemmy.mlhexagonhexbear2·4 months agoThe rumor is 1.76 trillion, or 8x220B (mixture of experts) to be specific: https://wandb.ai/byyoung3/ml-news/reports/AI-Expert-Speculates-on-GPT-4-Architecture---Vmlldzo0NzA0Nzg4 linkfedilink
340B is fucking huge, holy shit. How big is GPT-4?
The rumor is 1.76 trillion, or 8x220B (mixture of experts) to be specific: https://wandb.ai/byyoung3/ml-news/reports/AI-Expert-Speculates-on-GPT-4-Architecture---Vmlldzo0NzA0Nzg4