RSS BotMB to Hacker NewsEnglish · 7 hours agoFmllm: 4mb training data, 100mb model, Fibonacci embeddings, near-coherent. WTF?github.comexternal-linkmessage-square0fedilinkarrow-up13arrow-down10file-text
arrow-up13arrow-down1external-linkFmllm: 4mb training data, 100mb model, Fibonacci embeddings, near-coherent. WTF?github.comRSS BotMB to Hacker NewsEnglish · 7 hours agomessage-square0fedilinkfile-text
