Asynchronous Stochastic Gradient Descent with Decoupled Backpropagation and Layer-Wise Updates
- URL: http://arxiv.org/abs/2410.05985v1
- Date: Tue, 8 Oct 2024 12:32:36 GMT
- Title: Asynchronous Stochastic Gradient Descent with Decoupled Backpropagation and Layer-Wise Updates
- Authors: Cabrel Teguemne Fokam, Khaleelulla Khan Nazeer, Lukas König, David Kappel, Anand Subramoney,
- Abstract summary: One major shortcoming of backpropagation is the interlocking between the forward and backward phases of the algorithm.
We propose a method that parallelises SGD updates across the layers of a model by asynchronously updating them from multiple threads.
We show that this approach yields close to state-of-theart results while running up to 2.97x faster than Hogwild! scaled on multiple devices.
- Score: 1.9241821314180372
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: The increasing size of deep learning models has created the need for more efficient alternatives to the standard error backpropagation algorithm, that make better use of asynchronous, parallel and distributed computing. One major shortcoming of backpropagation is the interlocking between the forward phase of the algorithm, which computes a global loss, and the backward phase where the loss is backpropagated through all layers to compute the gradients, which are used to update the network parameters. To address this problem, we propose a method that parallelises SGD updates across the layers of a model by asynchronously updating them from multiple threads. Furthermore, since we observe that the forward pass is often much faster than the backward pass, we use separate threads for the forward and backward pass calculations, which allows us to use a higher ratio of forward to backward threads than the usual 1:1 ratio, reducing the overall staleness of the parameters. Thus, our approach performs asynchronous stochastic gradient descent using separate threads for the loss (forward) and gradient (backward) computations and performs layer-wise partial updates to parameters in a distributed way. We show that this approach yields close to state-of-the-art results while running up to 2.97x faster than Hogwild! scaled on multiple devices (Locally-Partitioned-Asynchronous-Parallel SGD). We theoretically prove the convergence of the algorithm using a novel theoretical framework based on stochastic differential equations and the drift diffusion process, by modeling the asynchronous parameter updates as a stochastic process.
Related papers
- BitPipe: Bidirectional Interleaved Pipeline Parallelism for Accelerating Large Models Training [5.7294516069851475]
BitPipe is a bidirectional interleaved pipeline parallelism for accelerating large models training.
We show that BitPipe improves the training throughput of GPT-style and BERT-style models by 1.05x-1.28x compared to the state-of-the-art synchronous approaches.
arXiv Detail & Related papers (2024-10-25T08:08:51Z) - AsyncDiff: Parallelizing Diffusion Models by Asynchronous Denoising [49.785626309848276]
AsyncDiff is a universal and plug-and-play acceleration scheme that enables model parallelism across multiple devices.
For the Stable Diffusion v2.1, AsyncDiff achieves a 2.7x speedup with negligible degradation and a 4.0x speedup with only a slight reduction of 0.38 in CLIP Score.
Our experiments also demonstrate that AsyncDiff can be readily applied to video diffusion models with encouraging performances.
arXiv Detail & Related papers (2024-06-11T03:09:37Z) - Inverse-Free Fast Natural Gradient Descent Method for Deep Learning [52.0693420699086]
We present a fast natural gradient descent (FNGD) method that only requires inversion during the first epoch.
FNGD exhibits similarities to the average sum in first-order methods, leading to the computational complexity of FNGD being comparable to that of first-order methods.
arXiv Detail & Related papers (2024-03-06T05:13:28Z) - Robust Fully-Asynchronous Methods for Distributed Training over General Architecture [11.480605289411807]
Perfect synchronization in distributed machine learning problems is inefficient and even impossible due to the existence of latency, package losses and stragglers.
We propose Fully-Asynchronous Gradient Tracking method (R-FAST), where each device performs local computation and communication at its own without any form of impact.
arXiv Detail & Related papers (2023-07-21T14:36:40Z) - Delay-adaptive step-sizes for asynchronous learning [8.272788656521415]
We show that it is possible to use learning rates that depend on the actual time-varying delays in the system.
For each of these methods, we demonstrate how delays can be measured on-line, present delay-adaptive step-size policies, and illustrate their theoretical and practical advantages over the state-of-the-art.
arXiv Detail & Related papers (2022-02-17T09:51:22Z) - Distributed stochastic optimization with large delays [59.95552973784946]
One of the most widely used methods for solving large-scale optimization problems is distributed asynchronous gradient descent (DASGD)
We show that DASGD converges to a global optimal implementation model under same delay assumptions.
arXiv Detail & Related papers (2021-07-06T21:59:49Z) - Gradient Coding with Dynamic Clustering for Straggler Mitigation [57.9123881133818]
GC-DC regulates the number of straggling workers in each cluster based on the straggler behavior in the previous iteration.
We numerically show that GC-DC provides significant improvements in the average completion time (of each iteration) with no increase in the communication load compared to the original GC scheme.
arXiv Detail & Related papers (2020-11-03T18:52:15Z) - Stochastic Optimization with Laggard Data Pipelines [65.20044914532221]
We show that "dataechoed" extensions of common optimization methods exhibit provable improvements over their synchronous counterparts.
Specifically, we show that in convex optimization with minibatches, data echoing affords speedups on the curvature-dominated part of the convergence rate, while maintaining the optimal statistical rate.
arXiv Detail & Related papers (2020-10-26T14:55:31Z) - Sparse Communication for Training Deep Networks [56.441077560085475]
Synchronous gradient descent (SGD) is the most common method used for distributed training of deep learning models.
In this algorithm, each worker shares its local gradients with others and updates the parameters using the average gradients of all workers.
We study several compression schemes and identify how three key parameters affect the performance.
arXiv Detail & Related papers (2020-09-19T17:28:11Z) - DaSGD: Squeezing SGD Parallelization Performance in Distributed Training
Using Delayed Averaging [4.652668321425679]
Minibatch gradient descent (SGD) algorithm requires workers to halt forward/back propagations.
DaSGD parallelizes SGD and forward/back propagations to hide 100% of the communication overhead.
arXiv Detail & Related papers (2020-05-31T05:43:50Z) - OD-SGD: One-step Delay Stochastic Gradient Descent for Distributed
Training [5.888925582071453]
We propose a novel technology named One-step Delay SGD (OD-SGD) to combine their strengths in the training process.
We evaluate our proposed algorithm on MNIST, CIFAR-10 and ImageNet datasets.
arXiv Detail & Related papers (2020-05-14T05:33:36Z)
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.