Showing papers by "Dan Alistarh published in 2022"

GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers

Abstract: Generative Pre-trained Transformer models, known as GPT or OPT, set themselves apart through breakthrough performance across complex language modelling tasks, but also by their extremely high computational and storage costs. Specifically, due to their massive size, even inference for large, highly-accurate GPT models may require multiple performant GPUs, which limits the usability of such models. While there is emerging work on relieving this pressure via model compression, the applicability and performance of existing compression techniques is limited by the scale and complexity of GPT models. In this paper, we address this challenge, and propose GPTQ, a new one-shot weight quantization method based on approximate second-order information, that is both highly-accurate and highly-efficient. Specifically, GPTQ can quantize GPT models with 175 billion parameters in approximately four GPU hours, reducing the bitwidth down to 3 or 4 bits per weight, with negligible accuracy degradation relative to the uncompressed baseline. Our method more than doubles the compression gains relative to previously-proposed one-shot quantization methods, preserving accuracy, allowing us for the first time to execute an 175 billion-parameter model inside a single GPU for generative inference. Moreover, we also show that our method can still provide reasonable accuracy in the extreme quantization regime, in which weights are quantized to 2-bit or even ternary quantization levels. We show experimentally that these improvements can be leveraged for end-to-end inference speedups over FP16, of around 3.25x when using high-end GPUs (NVIDIA A100) and 4.5x when using more cost-effective ones (NVIDIA A6000). The implementation is available at https://github.com/IST-DASLab/gptq.

The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models

Abstract: In this paper, we consider the problem of sparsifying BERT models, which are a key building block for natural language processing, in order to reduce their storage and computational cost. We introduce the Optimal BERT Surgeon (oBERT), an efficient and accurate pruning method based on approximate second-order information, which we show to yield state-of-the-art results in both stages of language tasks: pre-training and fine-tuning. Specifically, oBERT extends existing work on second-order pruning by allowing for pruning weight blocks, and is the first such method that is applicable at BERT scale. Second, we investigate compounding compression approaches to obtain highly compressed but accurate models for deployment on edge devices. These models significantly push boundaries of the current state-of-the-art sparse BERT models with respect to all metrics: model size, inference speed and task accuracy. For example, relative to the dense BERT-base, we obtain 10x model size compression with < 1% accuracy drop, 10x CPU-inference speedup with < 2% accuracy drop, and 29x CPU-inference speedup with < 7.5% accuracy drop. Our code, fully integrated with Transformers and SparseML, is available at https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT.

Optimal Brain Compression: A Framework for Accurate Post-Training Quantization and Pruning

Abstract: We consider the problem of model compression for deep neural networks (DNNs) in the challenging one-shot/post-training setting, in which we are given an accurate trained model, and must compress it without any retraining, based only on a small amount of calibration input data. This problem has become popular in view of the emerging software and hardware support for executing models compressed via pruning and/or quantization with speedup, and well-performing solutions have been proposed independently for both compression approaches. In this paper, we introduce a new compression framework which covers both weight pruning and quantization in a unified setting, is time- and space-efficient, and considerably improves upon the practical performance of existing post-training methods. At the technical level, our approach is based on an exact and efficient realization of the classical Optimal Brain Surgeon (OBS) framework of [LeCun, Denker, and Solla, 1990] extended to also cover weight quantization at the scale of modern DNNs. From the practical perspective, our experimental results show that it can improve significantly upon the compression-accuracy trade-offs of existing post-training methods, and that it can enable the accurate compound application of both pruning and quantization in a post-training setting.

SPDY: Accurate Pruning with Speedup Guarantees

Abstract: The recent focus on the efficiency of deep neural networks (DNNs) has led to significant work on model compression approaches, of which weight pruning is one of the most popular. At the same time, there is rapidly-growing computational support for efficiently executing the unstructured-sparse models obtained via pruning. Yet, most existing pruning methods minimize just the number of remaining weights, i.e. the size of the model, rather than optimizing for inference time. We address this gap by introducing SPDY, a new compression method which automatically determines layer-wise sparsity targets achieving a desired inference speedup on a given system, while minimizing accuracy loss. SPDY is composed of two new techniques: the first is an efficient dynamic programming algorithm for solving the speedup-constrained layer-wise compression problem assuming a set of given layer-wise sensitivity scores; the second is a local search procedure for determining accurate layer-wise sensitivity scores. Experiments across popular vision and language models show that SPDY guarantees speedups while recovering higher accuracy relative to existing strategies, both for one-shot and gradual pruning scenarios, and is compatible with most existing pruning approaches. We also extend our approach to the recently-proposed task of pruning with very little data, where we achieve the best known accuracy recovery when pruning to the GPU-supported 2:4 sparsity pattern.

GMP*: Well-Tuned Global Magnitude Pruning Can Outperform Most BERT-Pruning Methods

Abstract: We revisit the performance of the classic gradual magnitude pruning (GMP) baseline for large language models, focusing on the classic BERT benchmark on various popular tasks. Despite existing evidence in the literature that GMP performs poorly, we show that a simple and general variant, which we call GMP (cid:70) , can match and sometimes outperform more complex state-of-the-art methods. Our results provide a simple yet strong baseline for future work, highlight the importance of parameter tuning for baselines, and even improve the performance of the state-of-the-art second-order pruning method in this setting.

Near-Optimal Leader Election in Population Protocols on Graphs

Abstract: In the stochastic population protocol model, we are given a connected graph with n nodes, and in every time step, a scheduler samples an edge of the graph uniformly at random and the nodes connected by this edge interact. A fundamental task in this model is stable leader election, in which all nodes start in an identical state and the aim is to reach a configuration in which (1) exactly one node is elected as leader and (2) this node remains as the unique leader no matter what sequence of interactions follows. On cliques, the complexity of this problem has recently been settled: time-optimal protocols stabilize in Θ(n log n) expected steps using Θ(log log n) states, whereas protocols that use O(1) states require Θ(n2) expected steps. In this work, we investigate the complexity of stable leader election on general graphs. We provide the first non-trivial time lower bounds for leader election on general graphs, showing that, when moving beyond cliques, the complexity landscape of leader election becomes very diverse: the time required to elect a leader can range from O(1) to Θ(n3) expected steps. On the upper bound side, we first observe that there exists a protocol that is time-optimal on many graph families, but uses polynomially-many states. In contrast, we give a near-time-optimal protocol that uses only O(log2n) states that is at most a factor log n slower. Finally, we show that the constant-state protocol of Beauquier et al. [OPODIS 2013] is at most a factor n log n slower than the fast polynomial-state protocol. Moreover, among constant-state protocols, this protocol has near-optimal average case complexity on dense random graphs.

Fast and Scalable Channels in Kotlin Coroutines

Abstract: Asynchronous programming has gained significant popularity over the last decade: support for this programming pattern is available in many popular languages via libraries and native language implementations, typically in the form of coroutines or the async/await construct. Instead of programming via shared memory, this concept assumes implicit synchronization through message passing. The key data structure enabling such communication is the rendezvous channel. Roughly, a rendezvous channel is a blocking queue of size zero, so both send(e) and receive() operations wait for each other, performing a rendezvous when they meet. To optimize the message passing pattern, channels are usually equipped with a fixed-size buffer, so sends do not suspend and put elements into the buffer until its capacity is exceeded. This primitive is known as a buffered channel. This paper presents a fast and scalable algorithm for both rendezvous and buffered channels. Similarly to modern queues, our solution is based on an infinite array with two positional counters for send(e) and receive() operations, leveraging the unconditional Fetch-And-Add instruction to update them. Yet, the algorithm requires non-trivial modifications of this classic pattern, in order to support the full channel semantics, such as buffering and cancellation of waiting requests. We compare the performance of our solution to that of the Kotlin implementation, as well as against other academic proposals, showing up to 9.8× speedup. To showcase its expressiveness and performance, we also integrated the proposed algorithm into the standard Kotlin Coroutines library, replacing the previous channel implementations.

CrAM: A Compression-Aware Minimizer

Abstract: Deep neural networks (DNNs) often have to be compressed, via pruning and/or quantization, before they can be deployed in practical settings. In this work we propose a new compression-aware minimizer dubbed CrAM that modifies the optimization step in a principled way, in order to produce models whose local loss behavior is stable under compression operations such as pruning. Thus, dense models trained via CrAM should be compressible post-training, in a single step, without significant accuracy loss. Experimental results on standard benchmarks, such as residual networks for ImageNet classification and BERT models for language modelling, show that CrAM produces dense models that can be more accurate than the standard SGD/Adam-based baselines, but which are stable under weight pruning: specifically, we can prune models in one-shot to 70-80% sparsity with almost no accuracy loss, and to 90% with reasonable ($\sim 1\%$) accuracy loss, which is competitive with gradual compression methods. Additionally, CrAM can produce sparse models which perform well for transfer learning, and it also works for semi-structured 2:4 pruning patterns supported by GPU hardware. The code for reproducing the results is available at https://github.com/IST-DASLab/CrAM .

L-GreCo: An Efficient and General Framework for Layerwise-Adaptive Gradient Compression

Abstract: Data-parallel distributed training of deep neural networks (DNN) has gained very widespread adoption, but can still experience communication bottlenecks due to gradient transmission. To address this issue, entire families of lossy gradient compression mechanisms have been developed, including quantization, sparsification, and low-rank approximation, some of which are seeing significant practical adoption. Despite this progress, almost all known compression schemes apply compression uniformly across DNN layers, although layers are heterogeneous in terms of parameter count and their impact on model accuracy. In this work, we provide a general framework for adapting the degree of compression across the model’s layers dynamically during training, significantly improving the overall compression without sacrificing accuracy. Our framework, called L-GreCo, is based on an efficient adaptive algorithm, which automatically picks the optimal compression parameters for model layers guaranteeing the best compression ratio while respecting a theoretically-justified error constraint. Our extensive experimental study over image classification and language modeling tasks shows that L-GreCo is effective across all three compression families, and achieves up to 2.5× training speedup and up to 5× compression improvement over efficient implementations of standard approaches while recovering full accuracy. Moreover, we show that L-GreCo is complementary to existing adaptive algorithms improving their compression ratio by 50% and practical throughput by 66%. An anonymized implementation is available at https://github.com/LGrCo/L-GreCo.

oViT: An Accurate Second-Order Pruning Framework for Vision Transformers

Abstract: Models from the Vision Transformer (ViT) family have recently provided break-through results across image classification tasks such as ImageNet. Yet, they still face barriers to deployment, notably the fact that their accuracy can be severely impacted by compression techniques such as pruning. In this paper, we take a step towards addressing this issue by introducing Optimal ViT Surgeon (oViT) , a new state-of-the-art method for the weight sparsification of Vision Transformers (ViT) models. At the technical level, oViT introduces a new weight pruning algorithm which leverages second-order information, specifically adapted to be both highly-accurate and efficient in the context of ViTs. We complement this accurate one-shot pruner with an in-depth investigation of gradual pruning, augmentation, and recovery schedules for ViTs, which we show to be critical for successful ViT compression. We validate our method via extensive experiments on classical ViT and DeiT models, as well as on newer variants, such as XCiT, EfficientFormer and Swin. Moreover, our results are even relevant to recently-proposed highly-accurate ResNets. Our results show for the first time that ViT-family models can in fact be pruned to high sparsity levels (e.g. ≥ 75% ) with low impact on accuracy ( ≤ 1% relative drop), and that our approach outperforms prior methods by significant margins at high sparsities. In addition, we show that our method is compatible with structured pruning methods and quantization, and that it can lead to significant speedups on a sparsity-aware inference engine.

PathCAS: an efficient middle ground for concurrent search data structures

Abstract: To maximize the performance of concurrent data structures, researchers have often turned to highly complex fine-grained techniques, resulting in efficient and elegant algorithms, which can however be often difficult to understand and prove correct. While simpler techniques exist, such as transactional memory, they can have limited performance or portability relative to their fine-grained counterparts. Approaches at both ends of this complexity-performance spectrum have been extensively explored, but relatively less is known about the middle ground: approaches that are willing to sacrifice some performance for simplicity, while remaining competitive with state-of-the-art handcrafted designs. In this paper, we explore this middle ground, and present PathCAS, a primitive that combines ideas from multi-word CAS (KCAS) and transactional memory approaches, while carefully avoiding overhead. We show how PathCAS can be used to implement efficient search data structures relatively simply, using an internal binary search tree as an example, then extending this to an AVL tree. Our best implementations outperform many handcrafted search trees: in search-heavy workloads, it rivals the BCCO tree [5], the fastest known concurrent binary tree in terms of search performance [3]. Our results suggest that PathCAS can yield concurrent data structures that are relatively easy to build and prove correct, while offering surprisingly high performance.

Hybrid Decentralized Optimization: First- and Zeroth-Order Optimizers Can Be Jointly Leveraged For Faster Convergence

Abstract: first-order, we initiate the study of hybrid decentralized optimization, studying settings where nodes with zeroth-order and first-order optimization capabilities co-exist in a distributed system, and attempt to jointly solve an optimization task over some data distribution. We essentially show that, under reasonable parameter settings, such a system can not only withstand noisier zeroth-order agents, but can even benefit from integrating such agents into the optimization process, rather than ignoring their information. At the core of our approach is a new analysis of distributed optimization with noisy and possibly-biased gradient estimators, which may be of independent interest. Experimental results on standard optimization tasks confirm our analysis, showing that hybrid first-zeroth order optimization can be practical.

Scaling the Wild: Decentralizing Hogwild!-style Shared-memory SGD

Abstract: Powered by the simplicity of lock-free asynchrony, Hogwilld! is a go-to approach to parallelize SGD over a shared-memory setting. Despite its popularity and concomitant extensions, such as PASSM+ wherein concurrent processes update a shared model with partitioned gradients, scaling it to decentralized workers has surprisingly been relatively unexplored. To our knowledge, there is no convergence theory of such methods, nor systematic numerical comparisons evaluating speed-up. In this paper, we propose an algorithm incorporating decentralized distributed memory computing architecture with each node running multiprocessing parallel shared-memory SGD itself. Our scheme is based on the following algorithmic tools and features: (a) asynchronous local gradient updates on the shared-memory of workers, (b) partial backpropagation, and (c) non-blocking in-place averaging of the local models. We prove that our method guarantees ergodic convergence rates for non-convex objectives. On the practical side, we show that the proposed method exhibits improved throughput and competitive accuracy for standard image classification benchmarks on the CIFAR-10, CIFAR-100, and Imagenet datasets. Our code is available at https://github.com/bapi/LPP-SGD.

Cgx

Abstract: The ability to scale out training workloads has been one of the key performance enablers of deep learning. The main scaling approach is data-parallel GPU-based training, which has been boosted by hardware and software support for highly efficient point-to-point communication, and in particular via hardware bandwidth overprovisioning. Overprovisioning comes at a cost: there is an order of magnitude price difference between "cloud-grade" servers with such support, relative to their popular "consumer-grade" counterparts, although single server-grade and consumer-grade GPUs can have similar computational envelopes. In this paper, we show that the costly hardware overprovisioning approach can be supplanted via algorithmic and system design, and propose a framework called CGX, which provides efficient software support for compressed communication in ML applications, for both multi-GPU single-node training, as well as larger-scale multi-node training. CGX is based on two technical advances: \emph{At the system level}, it relies on a re-developed communication stack for ML frameworks, which provides flexible, highly-efficient support for compressed communication. \emph{At the application level}, it provides \emph{seamless, parameter-free} integration with popular frameworks, so that end-users do not have to modify training recipes, nor significant training code. This is complemented by a \emph{layer-wise adaptive compression} technique which dynamically balances compression gains with accuracy preservation. CGX integrates with popular ML frameworks, providing up to 3X speedups for multi-GPU nodes based on commodity hardware, and order-of-magnitude improvements in the multi-node setting, with negligible impact on accuracy.

Elastic Consistency

Abstract: Machine learning models can match or surpass humans on specialized tasks such as image classification [20, 14], speech recognition [37], or complex games [39]. One key tool behind this progress has been a family of optimization methods which fall under the umbrella term of stochastic gradient descent (SGD) [35], which are by and large the method of choice for training large-scale machine learning models.

L-GreCo: Layerwise-Adaptive Gradient Compression for Efficient and Accurate Deep Learning

Abstract: Data-parallel distributed training of deep neural networks (DNN) has gained very widespread adoption, but can still experience communication bottlenecks. To address this issue, entire families of compression mechanisms have been developed, including quantization, sparsification, and low-rank approximation, some of which are seeing significant practical adoption. Despite this progress, almost all known compression schemes apply compression uniformly across DNN layers, although layers are heterogeneous in terms of parameter count and their impact on model accuracy. In this work, we provide a general framework for adapting the degree of compression across the model's layers dynamically during training, improving the overall compression, while leading to substantial speedups, without sacrificing accuracy. Our framework, called L-GreCo, is based on an adaptive algorithm, which automatically picks the optimal compression parameters for model layers guaranteeing the best compression ratio while satisfying an error constraint. Extensive experiments over image classification and language modeling tasks shows that L-GreCo is effective across all existing families of compression methods, and achieves up to 2.5$\times$ training speedup and up to 5$\times$ compression improvement over efficient implementations of existing approaches, while recovering full accuracy. Moreover, L-GreCo is complementary to existing adaptive algorithms, improving their compression ratio by 50% and practical throughput by 66%.

Communication-Efficient Federated Learning With Data and Client Heterogeneity

Abstract: Federated Learning (FL) enables large-scale distributed training of machine learning models, while still allowing individual nodes to maintain data locally. However, executing FL at scale comes with inherent practical challenges: 1) heterogeneity of the local node data distributions, 2) heterogeneity of node computational speeds (asynchrony), but also 3) constraints in the amount of communication between the clients and the server. In this work, we present the first variant of the classic federated averaging (FedAvg) algorithm which, at the same time, supports data heterogeneity, partial client asynchrony, and communication compression. Our algorithm comes with a rigorous analysis showing that, in spite of these system relaxations, it can provide similar convergence to FedAvg in interesting parameter regimes. Experimental results in the rigorous LEAF benchmark on setups of up to $300$ nodes show that our algorithm ensures fast convergence for standard federated tasks, improving upon prior quantized and asynchronous approaches.

Distributed Computing Column 85 Elastic Consistency

Abstract: Overview. In this column, we attempt to cover a new and exciting area in distributed computing, that is, distributed and parallel machine learning. Together with my students Giorgi Nadiradze and Ilia Markov, I share an expository version of a recent conference paper [30] which attempts to provide a "consistency condition" for optimization problems which appear in distributed computing and machine learning applications.

PathCAS

Abstract: To maximize the performance of concurrent data structures, researchers have often turned to highly complex fine-grained techniques, resulting in efficient and elegant algorithms, which can however be often difficult to understand and prove correct. While simpler techniques exist, such as transactional memory, they can have limited performance or portability relative to their fine-grained counterparts. Approaches at both ends of this complexity-performance spectrum have been extensively explored, but relatively less is known about the middle ground: approaches that are willing to sacrifice some performance for simplicity, while remaining competitive with state-of-the-art handcrafted designs. In this paper, we explore this middle ground, and present PathCAS, a primitive that combines ideas from multi-word CAS (KCAS) and transactional memory approaches, while carefully avoiding overhead. We show how PathCAS can be used to implement efficient search data structures relatively simply, using an internal binary search tree as an example, then extending this to an AVL tree. Our best implementations outperform many handcrafted search trees: in search-heavy workloads, it rivals the BCCO tree [5], the fastest known concurrent binary tree in terms of search performance [3]. Our results suggest that PathCAS can yield concurrent data structures that are relatively easy to build and prove correct, while offering surprisingly high performance.

CAP: Correlation-Aware Pruning for Highly-Accurate Sparse Vision Models

Abstract: Driven by significant improvements in architectural design and training pipelines, computer vision has recently experienced dramatic progress in terms of accuracy on classic benchmarks such as ImageNet. These highly-accurate models are challenging to deploy, as they appear harder to compress using standard techniques such as pruning. We address this issue by introducing the Correlation Aware Pruner (CAP), a new unstructured pruning framework which significantly pushes the compressibility limits for state-of-the-art architectures. Our method is based on two technical advancements: a new theoretically-justified pruner, which can handle complex weight correlations accurately and efficiently during the pruning process itself, and an efficient finetuning procedure for post-compression recovery. We validate our approach via extensive experiments on several modern vision models such as Vision Transformers (ViT), modern CNNs, and ViT-CNN hybrids, showing for the first time that these can be pruned to high sparsity levels (e.g. $\geq 75$%) with low impact on accuracy ($\leq 1$% relative drop). Our approach is also compatible with structured pruning and quantization, and can lead to practical speedups of 1.5 to 2.4x without accuracy loss. To further showcase CAP's accuracy and scalability, we use it to show for the first time that extremely-accurate large vision models, trained via self-supervised techniques, can also be pruned to moderate sparsities, with negligible accuracy loss.

QuAFL: Federated Averaging Can Be Both Asynchronous and Communication-Efficient

Abstract: Federated Learning (FL) is an emerging paradigm to enable the large-scale distributed training of machine learning models, while still providing privacy guarantees. In this work, we jointly address two of the main practical challenges when scaling federated optimization to large node counts: the need for tight synchronization between the central authority and individual computing nodes, and the large communication cost of transmissions between the central server and clients. Specifically, we present a new variant of the classic federated averaging (FedAvg) algorithm, which supports both asynchronous communication and communication compression. We provide a new analysis technique showing that, in spite of these system relaxations, our algorithm essentially matches the best known bounds for FedAvg, under reasonable parameter settings. On the experimental side, we show that our algorithm ensures fast practical convergence for standard federated tasks.