lans_throwaway@alien.topB to LocalLLaMAEnglish · 3 years agoLook ahead decoding offers massive (~1.5x) speedup for inferenceplus-squarelmsys.orgexternal-linkmessage-square0linkfedilinkarrow-up11arrow-down10
arrow-up11arrow-down1external-linkLook ahead decoding offers massive (~1.5x) speedup for inferenceplus-squarelmsys.orglans_throwaway@alien.topB to LocalLLaMAEnglish · 3 years agomessage-square0linkfedilink
lans_throwaway@alien.topB to LocalLLaMAEnglish · 3 years agoLook ahead decoding offers massive (~1.5x) speedup for inferenceplus-squarelmsys.orgexternal-linkmessage-square4linkfedilinkarrow-up11arrow-down10
arrow-up11arrow-down1external-linkLook ahead decoding offers massive (~1.5x) speedup for inferenceplus-squarelmsys.orglans_throwaway@alien.topB to LocalLLaMAEnglish · 3 years agomessage-square4linkfedilink
minus-squarelans_throwaway@alien.topBtoLocalLLaMA•Dynamic LoRAs -- Crazy idea?linkfedilinkEnglisharrow-up1·3 years agoIsn’t that just Hypernetwork? It’s been done before, eg. for stable diffusion linkfedilink
Isn’t that just Hypernetwork? It’s been done before, eg. for stable diffusion