اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدSpAtten: Efficient Sparse Attention Architecture with Cascade Token and Head Pruning
62
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
The attention mechanism is becoming increasingly popular in Natural Language Processing (NLP) applications, showing superior performance than convolutional and recurrent architectures. However, general-purpose platforms such as CPUs and GPUs are inefficient when performing attention inference due to complicated data movement and low arithmetic intensity. Moreover, existing NN accelerators mainly focus on optimizing convolutional or recurrent models, and cannot efficiently support attention. In this paper, we present SpAtten, an efficient algorithm-architecture co-design that leverages token sparsity, head sparsity, and quantization opportunities to reduce the attention computation and memory access. Inspired by the high redundancy of human languages, we propose the novel cascade token pruning to prune away unimportant tokens in the sentence. We also propose cascade head pruning to remove unessential heads. Cascade pruning is fundamentally different from weight pruning since there is no trainable weight in the attention mechanism, and the pruned tokens and heads are selected on the fly. To efficiently support them on hardware, we design a novel top-k engine to rank token and head importance scores with high throughput. Furthermore, we propose progressive quantization that first fetches MSBs only and performs the computation; if the confidence is low, it fetches LSBs and recomputes the attention outputs, trading computation for memory reduction. Extensive experiments on 30 benchmarks show that, on average, SpAtten reduces DRAM access by 10.0x with no accuracy loss, and achieves 1.6x, 3.0x, 162x, 347x speedup, and 1,4x, 3.2x, 1193x, 4059x energy savings over A3 accelerator, MNNFast accelerator, TITAN Xp GPU, Xeon CPU, respectively.
قيم البحث
اقرأ أيضاً
Generalized Sparse Matrix-Matrix Multiplication (SpGEMM) is a ubiquitous task in various engineering and scientific applications. However, inner product based SpGENN introduces redundant input fetches for mismatched nonzero operands, while outer prod
uct based approach suffers from poor output locality due to numerous partial product matrices. Inefficiency in the reuse of either inputs or outputs data leads to extensive and expensive DRAM access. To address this problem, this paper proposes an efficient sparse matrix multiplication accelerator architecture, SpArch, which jointly optimizes the data locality for both input and output matrices. We first design a highly parallelized streaming-based merger to pipeline the multiply and merge stage of partial matrices so that partial matrices are merged on chip immediately after produced. We then propose a condensed matrix representation that reduces the number of partial matrices by three orders of magnitude and thus reduces DRAM access by 5.4x. We further develop a Huffman tree scheduler to improve the scalability of the merger for larger sparse matrices, which reduces the DRAM access by another 1.8x. We also resolve the increased input matrix read induced by the new representation using a row prefetcher with near-optimal buffer replacement policy, further reducing the DRAM access by 1.5x. Evaluated on 20 benchmarks, SpArch reduces the total DRAM access by 2.8x over previous state-of-the-art. On average, SpArch achieves 4x, 19x, 18x, 17x, 1285x speedup and 6x, 164x, 435x, 307x, 62x energy savings over OuterSPACE, MKL, cuSPARSE, CUSP, and ARM Armadillo, respectively.
Attention is sparse in vision transformers. We observe the final prediction in vision transformers is only based on a subset of most informative tokens, which is sufficient for accurate image recognition. Based on this observation, we propose a dynam
ic token sparsification framework to prune redundant tokens progressively and dynamically based on the input. Specifically, we devise a lightweight prediction module to estimate the importance score of each token given the current features. The module is added to different layers to prune redundant tokens hierarchically. To optimize the prediction module in an end-to-end manner, we propose an attention masking strategy to differentiably prune a token by blocking its interactions with other tokens. Benefiting from the nature of self-attention, the unstructured sparse tokens are still hardware friendly, which makes our framework easy to achieve actual speed-up. By hierarchically pruning 66% of the input tokens, our method greatly reduces 31%~37% FLOPs and improves the throughput by over 40% while the drop of accuracy is within 0.5% for various vision transformers. Equipped with the dynamic token sparsification framework, DynamicViT models can achieve very competitive complexity/accuracy trade-offs compared to state-of-the-art CNNs and vision transformers on ImageNet. Code is available at https://github.com/raoyongming/DynamicViT
Dot-product attention has wide applications in computer vision and natural language processing. However, its memory and computational costs grow quadratically with the input size. Such growth prohibits its application on high-resolution inputs. To re
medy this drawback, this paper proposes a novel efficient attention mechanism equivalent to dot-product attention but with substantially less memory and computational costs. Its resource efficiency allows more widespread and flexible integration of attention modules into a network, which leads to better accuracies. Empirical evaluations demonstrated the effectiveness of its advantages. Efficient attention modules brought significant performance boosts to object detectors and instance segmenters on MS-COCO 2017. Further, the resource efficiency democratizes attention to complex models, where high costs prohibit the use of dot-product attention. As an exemplar, a model with efficient attention achieved state-of-the-art accuracies for stereo depth estimation on the Scene Flow dataset. Code is available at https://github.com/cmsflash/efficient-attention.
A major challenge in deploying transformer models is their prohibitive inference cost, which quadratically scales with the input sequence length. This makes it especially difficult to use transformers for processing long sequences. To address this, w
e present a novel Learned Token Pruning (LTP) method that reduces redundant tokens as the data passes through the different layers of the transformer. In particular, LTP prunes tokens with an attention score below a threshold value, which is learned during training. Importantly, our threshold based method avoids algorithmically expensive operations such as top-k token selection which are used in prior token pruning methods, and also leads to structured pruning. We extensively test the performance of our approach on multiple GLUE tasks and show that our learned threshold based method consistently outperforms the prior state-of-the-art top-k token based method by up to ~2% higher accuracy with the same amount of FLOPs. Furthermore, our preliminary results show up to 1.4x and 1.9x throughput improvement on Tesla T4 GPU and Intel Haswell CPU, respectively, with less than 1% of accuracy drop (and up to 2.1x FLOPs reduction). Our code has been developed in PyTorch and has been open-sourced.
Convolutional neural networks (CNNs) have achieved great success in performing cognitive tasks. However, execution of CNNs requires a large amount of computing resources and generates heavy memory traffic, which imposes a severe challenge on computin
g system design. Through optimizing parallel executions and data reuse in convolution, systolic architecture demonstrates great advantages in accelerating CNN computations. However, regular internal data transmission path in traditional systolic architecture prevents the systolic architecture from completely leveraging the benefits introduced by neural network sparsity. Deployment of fine-grained sparsity on the existing systolic architectures is greatly hindered by the incurred computational overheads. In this work, we propose S2Engine $-$ a novel systolic architecture that can fully exploit the sparsity in CNNs with maximized data reuse. S2Engine transmits compressed data internally and allows each processing element to dynamically select an aligned data from the compressed dataflow in convolution. Compared to the naive systolic array, S2Engine achieves about $3.2times$ and about $3.0times$ improvements on speed and energy efficiency, respectively.
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
