Fine-tuning with Very Large Dropout
- URL: http://arxiv.org/abs/2403.00946v2
- Date: Tue, 22 Oct 2024 20:01:45 GMT
- Title: Fine-tuning with Very Large Dropout
- Authors: Jianyu Zhang, Léon Bottou,
- Abstract summary: It is impossible today to pretend that the practice of machine learning is compatible with the idea that training and testing data follow the same distribution.
Several authors have recently used ensemble techniques to show how scenarios involving multiple data distributions are best served by rich representations.
This contribution investigates the use of very high dropout rates instead of ensembles to obtain such rich representations.
- Score: 12.081782559668946
- License:
- Abstract: It is impossible today to pretend that the practice of machine learning is compatible with the idea that training and testing data follow the same distribution. Several authors have recently used ensemble techniques to show how scenarios involving multiple data distributions are best served by representations that are both richer than those obtained by regularizing for the best in-distribution performance, and richer than those obtained under the influence of the implicit sparsity bias of common stochastic gradient procedures. This contribution investigates the use of very high dropout rates instead of ensembles to obtain such rich representations. Although training a deep network from scratch using such dropout rates is virtually impossible, fine-tuning a large pre-trained model under such conditions is not only possible but also achieves out-of-distribution performances that exceed those of both ensembles and weight averaging methods such as model soups. This result has practical significance because the importance of the fine-tuning scenario has considerably grown in recent years. This result also provides interesting insights on the nature of rich representations and on the intrinsically linear nature of fine-tuning a large network using a comparatively small dataset.
Related papers
- ReFine: Boosting Time Series Prediction of Extreme Events by Reweighting and Fine-tuning [0.0]
Extreme events are of great importance since they represent impactive occurrences.
accurately predicting these extreme events is challenging due to their rarity and irregularity.
We propose two strategies, reweighting and fine-tuning, to tackle the challenge.
arXiv Detail & Related papers (2024-09-21T19:29:29Z) - Ask Your Distribution Shift if Pre-Training is Right for You [74.18516460467019]
In practice, fine-tuning a pre-trained model improves robustness significantly in some cases but not at all in others.
We focus on two possible failure modes of models under distribution shift: poor extrapolation and biases in the training data.
Our study suggests that, as a rule of thumb, pre-training can help mitigate poor extrapolation but not dataset biases.
arXiv Detail & Related papers (2024-02-29T23:46:28Z) - Learning Defect Prediction from Unrealistic Data [57.53586547895278]
Pretrained models of code have become popular choices for code understanding and generation tasks.
Such models tend to be large and require commensurate volumes of training data.
It has become popular to train models with far larger but less realistic datasets, such as functions with artificially injected bugs.
Models trained on such data tend to only perform well on similar data, while underperforming on real world programs.
arXiv Detail & Related papers (2023-11-02T01:51:43Z) - Uncertainty in Contrastive Learning: On the Predictability of Downstream
Performance [7.411571833582691]
We study whether the uncertainty of such a representation can be quantified for a single datapoint in a meaningful way.
We show that this goal can be achieved by directly estimating the distribution of the training data in the embedding space.
arXiv Detail & Related papers (2022-07-19T15:44:59Z) - Sliced-Wasserstein normalizing flows: beyond maximum likelihood training [12.91637880428221]
normalizing flows generally suffer from several shortcomings including their tendency to generate unrealistic data.
This paper proposes a new training paradigm based on a hybrid objective function combining the maximum likelihood principle (MLE) and a sliced-Wasserstein distance.
arXiv Detail & Related papers (2022-07-12T11:29:49Z) - Distributionally Robust Models with Parametric Likelihood Ratios [123.05074253513935]
Three simple ideas allow us to train models with DRO using a broader class of parametric likelihood ratios.
We find that models trained with the resulting parametric adversaries are consistently more robust to subpopulation shifts when compared to other DRO approaches.
arXiv Detail & Related papers (2022-04-13T12:43:12Z) - How Well Do Sparse Imagenet Models Transfer? [75.98123173154605]
Transfer learning is a classic paradigm by which models pretrained on large "upstream" datasets are adapted to yield good results on "downstream" datasets.
In this work, we perform an in-depth investigation of this phenomenon in the context of convolutional neural networks (CNNs) trained on the ImageNet dataset.
We show that sparse models can match or even outperform the transfer performance of dense models, even at high sparsities.
arXiv Detail & Related papers (2021-11-26T11:58:51Z) - An Effective Baseline for Robustness to Distributional Shift [5.627346969563955]
Refraining from confidently predicting when faced with categories of inputs different from those seen during training is an important requirement for the safe deployment of deep learning systems.
We present a simple, but highly effective approach to deal with out-of-distribution detection that uses the principle of abstention.
arXiv Detail & Related papers (2021-05-15T00:46:11Z) - Robust model training and generalisation with Studentising flows [22.757298187704745]
We discuss how these methods can be further improved based on insights from robust (in particular, resistant) statistics.
We propose to endow flow-based models with fat-tailed latent distributions as a simple drop-in replacement for the Gaussian distribution.
Experiments on several different datasets confirm the efficacy of the proposed approach.
arXiv Detail & Related papers (2020-06-11T16:47:01Z) - Mind the Trade-off: Debiasing NLU Models without Degrading the
In-distribution Performance [70.31427277842239]
We introduce a novel debiasing method called confidence regularization.
It discourages models from exploiting biases while enabling them to receive enough incentive to learn from all the training examples.
We evaluate our method on three NLU tasks and show that, in contrast to its predecessors, it improves the performance on out-of-distribution datasets.
arXiv Detail & Related papers (2020-05-01T11:22:55Z) - Robust and On-the-fly Dataset Denoising for Image Classification [72.10311040730815]
On-the-fly Data Denoising (ODD) is robust to mislabeled examples, while introducing almost zero computational overhead compared to standard training.
ODD is able to achieve state-of-the-art results on a wide range of datasets including real-world ones such as WebVision and Clothing1M.
arXiv Detail & Related papers (2020-03-24T03:59:26Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.