Sakeeb
sakeebrahman.bsky.social
Sakeeb
@sakeebrahman.bsky.social
Actively inferencing
Reposted by Sakeeb
BPE is a greedy method to find a tokeniser which maximises compression! Why don't we try to find properly optimal tokenisers instead? Well, it seems this is a pretty difficult—in fact, NP-complete—problem!🤯
New paper + @philipwitti.bsky.social
@gregorbachmann.bsky.social :) arxiv.org/abs/2412.15210
Tokenisation is NP-Complete
In this work, we prove the NP-completeness of two variants of tokenisation, defined as the problem of compressing a dataset to at most $δ$ symbols by either finding a vocabulary directly (direct token...
arxiv.org
December 20, 2024 at 2:04 PM
Reposted by Sakeeb
Super awkward to self-promote, but if you think I might be worth supporting or reading here is my Xmas discount special for all of December: (landing page to go premium) full disclosure Substack takes 10% and Stripe takes 3%.

This is my full-time job.
www.ai-supremacy.com/subscribe?co...
Subscribe to AI Supremacy
News at the intersection of Artificial Intelligence, technology and business including Op-Eds, research summaries, guest contributions and valuable info about A.I. startups. Click to read AI Supremacy...
www.ai-supremacy.com
December 4, 2024 at 4:56 AM
Reposted by Sakeeb
For anyone interested in fine-tuning or aligning LLMs, I’m running this free and open course called smol course. It’s not a big deal, it’s just smol.

🧵>>
December 3, 2024 at 9:21 AM
Reposted by Sakeeb
My deep learning course at the University of Geneva is available on-line. 1000+ slides, ~20h of screen-casts. Full of examples in PyTorch.

fleuret.org/dlc/

And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!)

fleuret.org/lbdl/
November 26, 2024 at 6:15 AM