News for hackers. Stuff that matters.
Re:Exponentially Faster Language Modelling by fgfm on Tuesday November 21, 2023 @06:59PM (#38372454)
This approach feels like pruning, but the speedup is considerably higher. Interestingly, I'm curious how this will play out on more recent transformer architectures though: I guess the speedup will be more important for the largest architectures, but even if we can get 2x or 10x speedup on Mistral/Zephyr, Orca 2 or OpenChat3.5, that would be a tremendous achievement!
Re:Exponentially Faster Language Modelling by webmaven on Wednesday November 22, 2023 @01:15AM (#38375368)
I'm curious as to how applicable this approach might be for text-to-image models like Stable Diffusion.
Re:Exponentially Faster Language Modelling by dang on Wednesday November 22, 2023 @01:00PM (#38382639)
Comments moved to https://news.ycombinator.com/item?id=38364084, which was posted earlier...