Daivik Dhar
0 followers · 233 views
on the atlas — 5
- Outperforming cuBLAS on H100: a Worklog1 savers
- Spring 2023 6.8210 Lecture 2: Nonlinear Dynamics - YouTube1 savers
- Ch. 21 - Imitation Learning2 savers
- [2410.05686] Deep Learning and Machine Learning with GPGPU and CUDA: Unlocking the Power of Parallel Computing1 savers
- 4.4 Value Iteration1 savers
highlights — 5
In this post, we’ll iteratively implement a CUDA kernel for matrix multiplication on latest generation1 NVIDIA hardware: H100.
Outperforming cuBLAS on H100: a Worklog13,486 views Streamed live on Feb 9, 2023 Underactuated Robotics, Spring 2023
Spring 2023 6.8210 Lecture 2: Nonlinear Dynamics - YouTubeAnother important aspect of the Diffusion Policy architecture is that it is predicting not just the instantaneous action, but a sequence of future actions. Even in the linear policy setting, we can see that this puts different pressure on the learned representations.
Ch. 21 - Imitation LearningGeneral Purpose Graphics Processing Unit (GPGPU) computing plays a transformative role in deep learning and machine learning by leveraging the computational advantages of parallel processing.
[2410.05686] Deep Learning and Machine Learning with GPGPU and CUDA: Unlocking the Power of Parallel ComputingAnother way of understanding value iteration is by reference to the Bellman optimality equation (4.1).
4.4 Value Iteration