people will finally understand why I moved from physics to AI
@thom_wolf
-
Training LLMs for Long Context Manipulation Tasks
By
–
Oh yes I see. Yes my hope is that it’s something solvable (i like on this topic eg Jina’s work) but we will need to train them more specifically on (long) context manipulation tasks. We’re lacking a bit of open benchmark/evals on this as well (simple manipulation without complex
-
Small AI Models: Strategic Use Beyond Knowledge Storage
By
–
You have to use these small models differently than the larger ones, typically not as knowledge ressource (they have limited storage capacity) for instance but more for creativity, reformulations, processing, etc. With external tools (eg search) and long context, they will become
-
Llama 3.2 achieves 10x improvement over previous generation
By
–
the new 1 billion parameters Llama model (version 3.2) is head-to-head with the 13 times larger version of one years ago (llama 13B version 2) on lmsys chatbot arena exciting to see such 10x improvements on challenging benchmark it's an amazing sign for small/local/open models
-
LLM Pleaser Tendency and Unconscious Role Matching
By
–
on the strong « pleaser » tendency of LLMs, matching the role you’re unconsciously asking them to play
-
LLama 3.2 1GB 2GB Versions Enable Ubiquitous Deployment
By
–
in case you missed it: LLama 3.2 in 1GB and 2GB sized version run it everywhere…
-
How to Create Podcasts with Google NotebookLM
By
–
By the way, here is how to do the NotebookLM stuff, just go to https://t.co/XldkWlK7J1 and do as I show below
— Thomas Wolf (@Thom_Wolf) 30 septembre 2024
A tip is to add some « production note dialogue » as free text if you want to control more what the dialogue contains (didn’t do it in the self-podcast example) https://t.co/bEuTezRg0q pic.twitter.com/S5UVqtDQGgBy the way, here is how to do the NotebookLM stuff, just go to https://
notebooklm.google.com and do as I show below A tip is to add some « production note dialogue » as free text if you want to control more what the dialogue contains (didn’t do it in the self-podcast example) -
NotebookLM Self-Care Hack: AI Podcast Praise Generator
By
–
Self-care life hack: if you feel a bit down/tired, paste the url of your website/linkedin/bio in Google's NotebookLM to get 8 min of realistically sounding deep congratulations for your life and achievements from a duo of podcast experts 😂 pic.twitter.com/k6krAgmMMd
— Thomas Wolf (@Thom_Wolf) 29 septembre 2024Self-care life hack: if you feel a bit down/tired, paste the url of your website/linkedin/bio in Google's NotebookLM to get 8 min of realistically sounding deep congratulations for your life and achievements from a duo of podcast experts
-
Open Multimodal LLM Releases Close Gap with Closed-Source Models
By
–
Celebrating a crazy month of "Open Multimodal LLM Releases"! If you haven't followed it several research labs have release impressively capable open multimodal LLMs in September 2024: Qwen-VL, Pixtral, LLama 3.2 and Molmo, seriously closing the gap with closed-source model
-
Building Transparent AI with Community Involvement Responsibly
By
–
Yes I see what you mean, the term "responsible AI" is quite overloaded A tweet won't make all the nuances but when I say "build AI responsibly", I means build AI transparently so I know what inside the AI I use and while involving the AI community as well as larger society in