The Lottery Ticket Hypothesis for Pre-trained BERT Networks

by   Tianlong Chen, et al.

In natural language processing (NLP), enormous pre-trained models like BERT have become the standard starting point for training on a range of downstream tasks, and similar trends are emerging in other areas of deep learning. In parallel, work on the lottery ticket hypothesis has shown that models for NLP and computer vision contain smaller matching subnetworks capable of training in isolation to full accuracy and transferring to other tasks. In this work, we combine these observations to assess whether such trainable, transferrable subnetworks exist in pre-trained BERT models. For a range of downstream tasks, we indeed find matching subnetworks at 40 subnetworks at (pre-trained) initialization, a deviation from prior NLP research where they emerge only after some amount of training. Subnetworks found on the masked language modeling task (the same task used to pre-train the model) transfer universally; those found on other tasks transfer in a limited fashion if at all. As large-scale pre-training becomes an increasingly central paradigm in deep learning, our results demonstrate that the main lottery ticket observations remain relevant in this context. Codes available at


page 7

page 8

page 14

page 15


The Lottery Tickets Hypothesis for Supervised and Self-supervised Pre-training in Computer Vision Models

The computer vision world has been re-gaining enthusiasm in various pre-...

DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter

As Transfer Learning from large-scale pre-trained models becomes more pr...

The Emergence of Essential Sparsity in Large Pre-trained Models: The Weights that Matter

Large pre-trained transformers are show-stealer in modern-day deep learn...

GAIA: A Transfer Learning System of Object Detection that Fits Your Needs

Transfer learning with pre-training on large-scale datasets has played a...

Playing Lottery Tickets with Vision and Language

Large-scale transformer-based pre-training has recently revolutionized v...

Towards Foundation Models for Scientific Machine Learning: Characterizing Scaling and Transfer Behavior

Pre-trained machine learning (ML) models have shown great performance fo...

ELLE: Efficient Lifelong Pre-training for Emerging Data

Current pre-trained language models (PLM) are typically trained with sta...

Code Repositories


[NeurIPS 2020] "The Lottery Ticket Hypothesis for Pre-trained BERT Networks", Tianlong Chen, Jonathan Frankle, Shiyu Chang, Sijia Liu, Yang Zhang, Zhangyang Wang, Michael Carbin

view repo

Please sign up or login with your details

Forgot password? Click here to reset