开启时会模糊预览图，关闭后正常显示
RT @neuranne: Tiny Experiments sold over 60K copies since March!! 🥹🙏❤️🔥🙌🥳 so grateful to everyone who bought the book and helped spread th…
hypercurious :) founder @ness_labs • neuroscientist @KingsIoPPN • author of Tiny Experiments • personal science, systematic curiosity, experimental thinking ꩜⋆✦
I just realised that the day we were attacked in ABJ was literally one year exactly to the date that this video was uploaded
Founder | Author | Speaker Building @beltstripe I'm Not The Man Of Your Dreams. Your Imagination Wasn't This Great.
The thing I'm doing on today with GPT-5 Pro is interesting because it's exactly the thing I was told was impossible: a deep, multi-step pair refactoring that pays down tech debt (somewhat accrued due to aggressive AI use) and towards re-using code due to totally new requirements
Opening portals to handheld VR at https://t.co/A2JMItorCV. Problems soluble, potential to improve invariant.
4/4 Physical Infrastructure for Cursor's Composer model training. They claim to have trained (and continue to train) across thousands of GPUs. They train models in low precision, and they use asynchronous RL (next tweet to explain what it is). Quote: "We built custom training infrastructure leveraging PyTorch and Ray to power asynchronous reinforcement learning at scale. We natively train our models at low precision by combining our MXFP8 MoE kernels with expert parallelism and hybrid sharded data parallelism, allowing us to scale training to thousands of NVIDIA GPUs with minimal communication cost. Additionally, training with MXFP8 allows us to deliver faster inference speeds without requiring post-training quantisation."
5/5 What is async RL that Customer Composer model training uses? It uses asynchronous execution at multiple levels to avoid waiting on slow operations e.g. a long roll-out generation. As you know, for a given problem, in RL like GRPO we generate multiple trajectorier. However, some trajectories can take too long to complete. So, once they have enough trajectories, they run the training. Partial samples/roll-outs are resumed later with updated model. This causes a situation where some tokens are generated by the old model/policy and some by new. However, this is acceptable. If you want to understand more about Async RL, please read APRIL - a project for Async RL.
i have a bunch of markov chain monte carlo stuff lying around from grad school i can throw at this, just have to port it from matlab
back in the prehistory of 3d computer vision (2016) we would use probabilistic / ebm models to fit shape templates to street scenes
RT @tibo_maker: SuperX was an "obvious product" my thought: "if I grew Tweet Hunter to $250k MRR, it should be easy with SuperX" yet,…
Built Tweet Hunter, Taplio (sold $8m) Growing https://t.co/OyNJ8ZUyOh - https://t.co/jS9GQJ5Ps8 - https://t.co/EFUcKeBbpU - https://t.co/JkVOl1O0S1 - https://t.co/KG9PgxJabg Sharing weekly tips about growth: https://t.co/ereQodN3Ov