The paper parallelizes HMM inference for efficient long-term computations.
problem Efficiently computing inference in long-term hidden Markov models.
method Parallelization using associative elements and operators for sum-product and max-product algorithms.
result The proposed parallel algorithms are computationally efficient for long time horizons.
We propose a new integrated method of exploiting model, batch and domain parallelism for the training of deep neural networks (DNNs) on large distributed-memory computers using minibatch stochastic gradient descent (SGD). Our goal is to find an efficient parallelization strategy for a fixed batch size using P process…
Linear algebra approach for parallel deep learning models.
problem Training large DNNs in distributed environments.
method Linear algebraic approach to model parallelism.
result Manual development of backward operators for gradient-based training.
Cyclic Data Parallelism reduces memory usage and balances gradient communications.
problem Training large deep learning models requires efficient parallelism to scale.
method Cyclic Data Parallelism shifts micro-batches from simultaneous to sequential execution, balancing memory and gradient communications.
result Cyclic Data Parallelism reduces total memory usage and balances gradient communications.
Parallelizes autoregressive generation using VSSM.
problem Autoregressive models' inability to parallelize generation.
method Variational SSM (VSSM) with parallelizable sampling and decoding.
result Parallel generation possible with VSSM.
NeLLoC improves image compression with parallel decoding.
problem Image compression with OOD generalization.
method Local autoregressive model with parallel decoding.
result Significant gains in compression runtime.
Improves parallel deep model performance by restructuring and pruning.
problem Latency in parallel deep model execution due to interdependency among sub-models.
method Layer-wise model restructuring and pruning, using ℓ0 optimization and Munkres assignment algorithm. result Significantly improves efficiency of distributed inference in terms of communication and computational complexity.
In real world industrial applications of topic modeling, the ability to capture gigantic conceptual space by learning an ultra-high dimensional topical representation, i.e., the so-called "big model", is becoming the next desideratum after enthusiasms on "big data", especially for fine-grained downstream tasks such as …
This paper improves parallel belief propagation for scalable machine learning.
problem Efficient parallelization of belief propagation for large-scale machine learning tasks.
method Use of scalable relaxed schedulers to parallelize belief propagation.
result Our approach outperforms previous methods in scalability and convergence time.
Adaptive batch size schedules improve language model training efficiency and generalization.
problem Dilemma of choosing batch sizes in large-scale model training.
method General-purpose adaptive batch size schedules compatible with data and model parallelism.
result Adaptive batch size schedules outperform constant batch sizes and heuristic warmup schedules.
We introduce a new sampling method for large language models that balances diversity and parallelism.
problem Balancing diversity and parallelism in decoding for large language models.
method Arithmetic sampling framework compatible with various sampling variations.
result Improves estimation of expected BLEU score reward and reduces the gap with beam search.
Mozilla Research is developing Servo, a parallel web browser engine, to exploit the benefits of parallelism and concurrency in the web rendering pipeline. Parallelization results in improved performance for pinterest.com but not for google.com. This is because the workload of a browser is dependent on the web page it i…
Serial problems can't be efficiently parallelized, affecting machine learning models.
problem Inefficiency of parallelization in inherently serial problems.
method Formalized distinction in complexity theory, demonstrated with diffusion models.
result Diffusion models cannot solve inherently serial problems.
The complex and computationally expensive nature of landscape evolution models pose significant challenges in the inference and optimisation of unknown parameters. Bayesian inference provides a methodology for estimation and uncertainty quantification of unknown model parameters. In our previous work, we developed para…
We introduce a new embarrassingly parallel parameter learning algorithm for Markov random fields with untied parameters which is efficient for a large class of practical models. Our algorithm parallelizes naturally over cliques and, for graphs of bounded degree, its complexity is linear in the number of cliques. Unlike…
Paper proposes DCT for efficient hybrid parallel training of large recommendation models.
problem Training large recommendation models at scale with efficient communication.
method Dynamic Communication Thresholding (DCT) for both Data Parallelism and Model Parallelism.
result Reduces communication by 100x and 20x during DP and MP, respectively, improving training time by 37%.
This research optimizes energy consumption forecasting in Puno using parallel computing and ARIMA models.
problem Improving energy consumption forecasting accuracy and efficiency in Puno.
method Parallel computing and ARIMA models for forecasting energy consumption.
result Notable improvements in computational efficiency and data processing capabilities.
PETRA enables parallel training of deep models with reversible architectures.
problem Challenges in parallelizing deep model training.
method Introduces PETRA, a novel approach for parallelizing gradient computations in reversible architectures.
result Achieves competitive accuracies on CIFAR-10, ImageNet32, and ImageNet using ResNet models.
Embarrassingly (communication-free) parallel Markov chain Monte Carlo (MCMC) methods are commonly used in learning graphical models. However, MCMC cannot be directly applied in learning topic models because of the quasi-ergodicity problem caused by multimodal distribution of topics. In this paper, we develop an embarra…
We develop a parallel variational inference (VI) procedure for use in data-distributed settings, where each machine only has access to a subset of data and runs VI independently, without communicating with other machines. This type of "embarrassingly parallel" procedure has recently been developed for MCMC inference al…
Batch-splitting (data-parallelism) is the dominant distributed Deep Neural Network (DNN) training strategy, due to its universal applicability and its amenability to Single-Program-Multiple-Data (SPMD) programming. However, batch-splitting suffers from problems including the inability to train very large models (due to…
Deploying deep learning (DL) models across multiple compute devices to train large and complex models continues to grow in importance because of the demand for faster and more frequent training. Data parallelism (DP) is the most widely used parallelization strategy, but as the number of devices in data parallel trainin…
Study the effects of data parallelism and sparsity on neural network training.
problem Understanding the effects of data parallelism and sparsity on neural network training.
method Conducted extensive experiments and developed a theoretical analysis.
result Found a general scaling trend between batch size and number of training steps to convergence for the effect of data parallelism, and difficulty of training under sparsity.
Two parallel samplers enhance image quality in limited denoising steps.
problem Limited denoising steps in diffusion models reduce image quality.
method Two parallel samplers denoise at successive times, integrating their information.
result Two parallel samplers improve image quality compared to a single sampler.
Speeds up deep neural networks training by 10x using GPU concurrency.
problem Training deep residual neural networks efficiently.
method Layer-wise parallel training with GPU concurrency and Nonlinear Multigrid.
result 10.2x speedup over traditional techniques.
To scale non-parametric extensions of probabilistic topic models such as Latent Dirichlet allocation to larger data sets, practitioners rely increasingly on parallel and distributed systems. In this work, we study data-parallel training for the hierarchical Dirichlet process (HDP) topic model. Based upon a representati…
MixML unifies analysis of weakly consistent parallel learning.
problem Lack of insight into how communication structure affects convergence in parallel learning.
method Proposes MixML framework for analyzing convergence of weakly consistent parallel machine learning.
result Shows dependency of convergence on mixing time tmix.
New method uses higher-order Langevin dynamics for efficient parallel sampling.
problem Efficient parallel sampling from high-dimensional log-concave distributions.
method Combines higher-order Langevin dynamics with blockwise Lagrange polynomial interpolation.
result Reduces the number of parallel points required for a target accuracy.
We introduce topological parallelisms of oriented lines (briefly called oriented parallelisms). Every topological parallelism (of lines) on PG(3,R) gives rise to a parallelism of oriented lines, but we show that even the most homogeneous parallelisms of oriented lines other than the Clifford parallelism do not necessar…
We suggest a general oracle-based framework that captures different parallel stochastic optimization settings described by a dependency graph, and derive generic lower bounds in terms of this graph. We then use the framework and derive lower bounds for several specific parallel optimization settings, including delayed …
Parallel score matching accelerates DPM training and improves density estimation.
problem Extended training periods and limited modeling flexibility in DPMs.
method Partitioning the learning task into independent time sub-intervals and modeling the score at each time point separately.
result Significant acceleration of training process and improved density estimation performance.
Paper improves financial trading models using GPU parallelism.
problem Challenges in policy instability and sampling bottlenecks in reinforcement learning for financial tasks.
method Revisits ensemble methods with massively parallel simulations on GPUs.
result Significantly improved computational efficiency and robustness of financial decision-making strategies.
Parallelizes MCTS for continuous domains using leaf and root parallelization.
problem Solving challenging tasks in continuous domains using MCTS.
method Extends existing parallelization strategies to continuous domains, focusing on leaf and root parallelization.
result Proposes two final selection strategies for continuous states in root parallelization.
TensorOpt finds optimal parallelization strategies for DNN training.
problem Finding efficient parallelization strategies for DNN training.
method TensorOpt uses an algorithm (FT) to search for an optimal set of parallelization strategies considering multiple objectives.
result TensorOpt provides accurate runtime cost estimation and adapts to resource availability.
Parallel unlearning framework for inherited models reduces computational overhead.
problem Challenges in unlearning complex, evolving model networks.
method Chronologically Directed Acyclic Graph (DAG) and Fisher Inheritance Unlearning (FIUn) method.
result Significant reduction in computational overhead and efficient parallel unlearning.
A deep-learning inference accelerator is synthesized from a C-language software program parallelized with Pthreads. The software implementation uses the well-known producer/consumer model with parallel threads interconnected by FIFO queues. The LegUp high-level synthesis (HLS) tool synthesizes threads into parallel FPG…
Proposes a method to improve SLMC for multimodal distributions.
problem Difficulty of applying SLMC to multimodal distributions.
method Parallel adaptive annealing with VAE-SLMC.
result Can proficiently obtain accurate samples from multimodal distributions.
Predictability enables efficient parallelization of nonlinear models.
problem Understanding which nonlinear state space models can be efficiently parallelized.
method Established a relationship between system dynamics and optimization problem conditioning, quantified by the largest Lyapunov exponent.
result Predictable systems can be evaluated in O((logT)2) time, improving over conventional sequential approaches. TSSM splits neural networks for parallel training with minimal accuracy loss.
problem Accuracy degradation in parallel training of deep neural networks.
method TSSM reformulates alternating minimization to achieve parallelism with minimal accuracy loss.
result TSSM achieves significant speedup without accuracy loss on multiple datasets.
Gaussian processes (GP) are Bayesian non-parametric models that are widely used for probabilistic regression. Unfortunately, it cannot scale well with large data nor perform real-time predictions due to its cubic time cost in the data size. This paper presents two parallel GP regression methods that exploit low-rank co…
Gaussian processes (GP) are Bayesian non-parametric models that are widely used for probabilistic regression. Unfortunately, it cannot scale well with large data nor perform real-time predictions due to its cubic time cost in the data size. This paper presents two parallel GP regression methods that exploit low-rank co…
HPSGD speeds up DNN training by paralleling data sync with local training.
problem Low cluster utilization in distributed deep neural network training.
method Hierarchical Parallel SGD (HPSGD) with improved model updating for stale gradients.
result Significantly boosts distributed DNN training and reduces stale gradients.
New compact Weyl-parallel manifolds discovered in all dimensions n≥5.
problem Finding compact Weyl-parallel manifolds in all metric signatures and dimensions.
method Diffeomorphic to torus bundles over the circle, constructed from quotient-manifolds of model manifolds with discrete isometry groups.
result Existence of compact Weyl-parallel manifolds in all indefinite metric signatures in dimensions n≥5.
AgEBO-Tabular combines NAS and hyperparameter tuning for fast, high-performing tabular models.
problem Developing high-performing predictive models for large tabular data sets is challenging.
method Combines aging evolution NAS and asynchronous Bayesian optimization for hyperparameter tuning in data-parallel training.
result Automatically discovered neural network models outperform state-of-the-art AutoML ensembles in inference speed by two orders of magnitude.
This work proposes an efficient autoregressive model for text generation.
problem The challenge of generating high-quality text with autoregressive models.
method Introduces a cascaded decoding approach using Markov transformers to achieve sub-linear parallel time generation.
result Shows competitive accuracy/speed tradeoff compared to existing methods on five machine translation datasets.
We introduce a novel approach for parallelizing MCMC inference in models with spatially determined conditional independence relationships, for which existing techniques exploiting graphical model structure are not applicable. Our approach is motivated by a model of seismic events and signals, where events detected in d…
When training large machine learning models with many variables or parameters, a single machine is often inadequate since the model may be too large to fit in memory, while training can take a long time even with stochastic updates. A natural recourse is to turn to distributed cluster computing, in order to harness add…
Parallel neural network training yields better long-term prediction accuracy.
problem Choosing the right training strategy for neural networks in dynamical systems.
method Comparison of parallel and series-parallel training strategies on five neural network architectures and two examples.
result Parallel training consistently outperforms series-parallel training in long-term prediction accuracy.