Hacker News new | ask | show | jobs
user: Mougatine
created: 2018-01-02
karma: 296

RS @ DeepMind.

Doing distributed stuff, such as DiLoCo and DiPaCo

submissions:

Fault Tolerant Llama training
66 points | 14 comments
MuLoCo: Muon is a practical inner optimizer for DiLoCo
2 points | 0 comments
0 points | 0 comments
OpenDiLoCo: Open-Source Framework for Distributed Low-Communication Training
4 points | 0 comments
0 points | 0 comments
0 points | 0 comments
0 points | 0 comments
0 points | 0 comments
0 points | 0 comments
0 points | 0 comments
0 points | 0 comments
0 points | 0 comments
0 points | 0 comments
0 points | 0 comments
Show HN: Deep Learning for Computer Vision course with colabs and Anki cards
12 points | 2 comments
0 points | 0 comments
0 points | 0 comments
Continual Learning at CVPR 2020
1 points | 0 comments
Operation Red Falcon (2015)
1 points | 0 comments
Lifelong Learning for Deep Neural Networks (2019)
2 points | 0 comments
0 points | 0 comments
0 points | 0 comments
Seeing Is Not Necessarily Believing: Limitations of GANs for Data Augmentation
2 points | 0 comments
Cars detection from satellite imagery with RetinaNet
2 points | 0 comments
Human or Company
1 points | 0 comments
3 Small but Powerful Convolutional Networks
4 points | 0 comments
An Explanation of Densely Connected Convolutional Networks
1 points | 0 comments
Amazon launches an Android app in India called “Internet”
1 points | 0 comments
Summary of “Deep Learning Scaling Is Predictable, Empirically”
1 points | 0 comments