Ted - 🥖/acc
@ted_engineer
Followers
127
Following
646
Media
37
Statuses
593
🇫🇷 / 25yo / dad / deep learning chef @duonlabshq
Savoie
Joined August 2023
I’m beyond excited to work on Apogée! 🚀. Let’s take the LLM revolution beyond text and into new frontiers!. #OpenSource #TimeSeries #ScalingLaws #Speedrunning.
Introducing Apogée. 🚀. We’re running the first large-scale crypto market scaling law experiment to answer the question: Can bigger models, trained on more candlestick data, extract more bits of the future ?
1
0
4
RT @__tinygrad__: This is tinygrad's description of the tensor cores of all the major GPUs. No per GPU dialects, just a spec for what they….
0
41
0
RT @Dorialexander: Model training used to be super conservative but it really fells we are at a breaking point. Grok scaling RL to pretrain….
0
24
0
RT @_albertgu: Tokenization is just a special case of "chunking" - building low-level data into high-level abstractions - which is in turn….
0
171
0
RT @fchollet: In order to supervise an automation tool (or another person!) effectively, you need to be able to do the same job yourself.….
0
168
0
RT @Thom_Wolf: We’re releasing the top 3B model out there. SOTA performances. It has dual mode reasoning (with or without think). Extended….
0
81
0
RT @_albertgu: I converted one of my favorite talks I've given over the past year into a blog post. "On the Tradeoffs of SSMs and Transfor….
0
112
0
RT @justalexoki: my honest reaction of switching to an iphone after having used android for a decade
0
139
0
the most common error of a smart.ml engineer is to train a model on a dataset that.should not exist.
0
0
1
RT @BetterCallMedhi: On parle beaucoup d’IA plus ou moins à juste titre (les tendances entraînent du BS), cependant je pense que le milieu….
0
12
0
RT @kalomaze: lr=1.0, clip=1e-10 → 99.13% of parameters unchanged compared to the base model. (orange run).lr=0.01, clip=1e-8 → 95.21% of p….
0
2
0
RT @kalomaze: hey guys did you know SWEBench is like ~70% one single repository and that one repository is Django.
0
41
0
RT @giffmana: Big fat metal boxes cannot & will not ever be able to float in the sky. Metal boxes will be able to roll on the ground when p….
0
60
0
RT @xueqinjiang: In this talk, I use game theory to examine how the US-Iran conflict might play out:
0
69
0
RT @leloykun: Fast, Numerically Stable, and Auto-Differentiable Spectral Clipping via Newton-Schulz Iteration. Hi all, I'm bacc. I have a l….
0
37
0