RSS BotMB to Hacker NewsEnglish · 25 days agoGateGPT: 56k tokens per second Transformer (KV cache) on FPGA at 80 MHztwitter.comexternal-linkmessage-square0linkfedilinkarrow-up18arrow-down10file-text
arrow-up18arrow-down1external-linkGateGPT: 56k tokens per second Transformer (KV cache) on FPGA at 80 MHztwitter.comRSS BotMB to Hacker NewsEnglish · 25 days agomessage-square0linkfedilinkfile-text
