activity
20182025
most citedThe large learning rate phase of deep learning: the catapult mechanism

58 citations · 123 across the 4 of their papers we have counts for

collaborators

12 papers

cs.CL20251 cited

Multilingual Machine Translation with Open Large Language Models at Practical Scale: An Empirical Study

Menglong Cui, Pengzhi Gao, Wei Liu +2

Large language models (LLMs) have shown continuously improving multilingual capabilities, and even small-scale open-source models have demonstrated rapid performance enhancement. I…

cs.LG202248 cited

Scaling Up Models and Data with and

Adam Roberts, Hyung Won Chung, Anselm Levskaya +40

Recent neural network-based language models have benefited greatly from scaling up the size of training datasets and the number of parameters in the models themselves. Scaling can…

cs.LG202116 cited

How to decay your learning rate

Aitor Lewkowycz

Complex learning rate schedules have become an integral part of deep learning. We find empirically that common fine-tuned schedules decay the learning rate after the weight norm bo…

hep-th2020

Gravitational path integral from the deformation

Alexandre Belin, Aitor Lewkowycz, Gabor Sarosi

We study a deformation of large conformal field theories, a higher dimensional generalization of the deformation. The deformed partition function satisfies a fl…

stat.ML2020

On the training dynamics of deep networks with regularization

Aitor Lewkowycz, Guy Gur-Ari

We study the role of regularization in deep learning, and uncover simple relations between the performance of the model, the coefficient, the learning rate, and the num…

stat.ML202058 cited

The large learning rate phase of deep learning: the catapult mechanism

Aitor Lewkowycz, Yasaman Bahri, Ethan Dyer +2

The choice of initial learning rate can have a profound effect on the performance of deep networks. We present a class of neural networks with solvable training dynamics, and confi…