↑ 点击蓝字关注极市平台

来源丨NewBeeNLP

编辑｜极市平台

极市导读

如何对attention进行高效改进？本文盘点了相关论文，并梳理出它们的引用量、代码实现、算法复杂度和关键点，方便对比使用。

前几天逛github刷到一个awesome-fast-attention大列表，整理了一系列关于attention的高效改进文章，包括论文、引用量、源码实现、算法复杂度以及关键亮点。

Github地址：

https://github.com/Separius/awesome-fast-attention

Efficient Attention

Paper (引用量)	源码实现	Main Idea
Generating Wikipedia by Summarizing Long Sequences[1] (208)	memory-compressed-attention[2]	compresses key and value + blocked attention
CBAM: Convolutional Block Attention Module[3] (677)	attention-module[4]	combines the SE attention with a per pixel(local) weight
CCNet: Criss-Cross Attention for Semantic Segmentation[5] (149)	CCNet[6]	each pixel attends to its row and column simultaneously
Efficient Attention: Attention with Linear Complexities[7] (2)	efficient-attention[8]	Softmax(Q)(Softmax(K^T)V)
Star-Transformer[9] (24)	fastNLP[10]	uses a relay(global) node and attends to/from that node
Generating Long Sequences with Sparse Transformers[11] (139)	torch-blocksparse[12]	sparse block based attention
GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond[13] (96)	GCNet[14]	squeeze and excitation with an attention pooling (instead of a GAP)
SCRAM: Spatially Coherent Randomized Attention Maps[15] (1)	-	uses PatchMatch to find close keys
Interlaced Sparse Self-Attention for Semantic Segmentation[16] (13)	IN_PAPER	combination of a short length and then long range(dilated) attention
Permutohedral Attention Module for Efficient Non-Local Neural Networks[17] (2)	Permutohedral_attention_module[18]	uses permutohedral lattice approximation algorithm to approximate the attention output
Large Memory Layers with Product Keys[19] (28)	XLM[20]	search for nearest neighbor keys
Expectation-Maximization Attention Networks for Semantic Segmentation[21] (38)	EMANet[22]	applys expectation maximization to cluster keys into k clusters
Compressive Transformers for Long-Range Sequence Modelling[23] (20)	compressive-transformer-pytorch[24]	compresses distant tokens instead of just stop_grad() ing them, more efficient version of transformerXL
BP-Transformer: Modelling Long-Range Context via Binary Partitioning[25] (8)	BPT[26]	attends to distant tokens coarsely and attends to close tokens in a more fine-grained manner
Axial Attention in Multidimensional Transformers[27] (5)	axial-attention[28]	apply attention on each axis separately
Reformer: The Efficient Transformer[29] (69)	trax[30]	uses LSH to find close keys
Transformer on a Diet[31] (2)	transformer-on-diet[32]	dilated transformer like wavenet
Sparse Sinkhorn Attention[33] (4)	sinkhorn-transformer[34]	uses a cost matrix to limit attention between buckets
SAC: Accelerating and Structuring Self-Attention via Sparse Adaptive Connection[35] (1)	-	learns the q, k connections == dynamically creates a sparse attention matrix
Efficient Content-Based Sparse Attention with Routing Transformers[36] (11)	routing-transformer[37]	computes attention with same-cluster tokens (computed by online k-means)
Longformer: The Long-Document Transformer[38] (15)	longformer[39]	global + blocked attention
Neural Architecture Search for Lightweight Non-Local Networks[40] (2)	AutoNL[41]	computes Q(KV) and also down samples q, k, v both in spatial and channel dimensions
ETC: Encoding Long and Structured Data in Transformers[42] (2)	-	combines global attention (star transformer with multiple global tokens) with local attention
Multi-scale Transformer Language Models[43] (1)	IN_PAPER	UNet like + retina attetion is something close to BP-Transformer
Synthesizer: Rethinking Self-Attention in Transformer Models[44] (5)	-	does not compute pairwise interactions
Jukebox: A Generative Model for Music[45] (9)	jukebox[46]	better attention patterns from Sparse Transformer
GMAT: Global Memory Augmentation for Transformers[47] (0)	gmat[48]	adds global tokens
Masked Language Modeling for Proteins via Linearly Scalable Long-Context Transformers[49] (0)	google-research[50]	calculate an unbiased stochastic approximation of the attention matrix
Hand-crafted Attention is All You Need? A Study of Attention on Self-supervised Audio Transformer[51] (0)	-	does not compute pairwise interactions and uses fixed mask patters
Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention[52] (1)	fast-transformers[53]	uses phi(q)(phi(k)v) and also improves the sequential sampling step
Linformer: Self-Attention with Linear Complexity[54] (3)	linformer-pytorch[55]	project key and value from nd
Real-time Semantic Segmentation with Fast Attention[56] (0)	-	l2_norm(q)(l2_norm(k)v)
Fast Transformers with Clustered Attention[57] (0)	fast-transformers[58]	groups queries together with LSH
Big Bird: Transformers for Longer Sequences[59] (0)	-	ETC with random connections

参考资料

[1] Generating Wikipedia by Summarizing Long Sequences: https://arxiv.org/abs/1801.10198v1

[2]memory-compressed-attention: https://github.com/lucidrains/memory-compressed-attention

[3] CBAM: Convolutional Block Attention Module: https://arxiv.org/abs/1807.06521v2

[4] attention-module: https://github.com/Jongchan/attention-module

[5] CCNet: Criss-Cross Attention for Semantic Segmentation: https://arxiv.org/abs/1811.11721v2

[6] CCNet: https://github.com/speedinghzl/CCNet

[7] Efficient Attention: Attention with Linear Complexities: https://arxiv.org/abs/1812.01243v8

[8] Efficient-attention: https://github.com/cmsflash/efficient-attention

[9] Star-Transformer: https://arxiv.org/abs/1902.09113v2

[10] fastNLP: https://github.com/fastnlp/fastNLP/blob/master/fastNLP/modules/encoder/star_transformer.py

[11] Generating Long Sequences with Sparse Transformers: https://arxiv.org/abs/1904.10509v1

[12] torch-blocksparse: https://github.com/ptillet/torch-blocksparse

[13] GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond: https://arxiv.org/abs/1904.11492v1

[14] GCNet: https://github.com/xvjiarui/GCNet

[15] SCRAM: Spatially Coherent Randomized Attention Maps: https://arxiv.org/abs/1905.10308v1

[16] Interlaced Sparse Self-Attention for Semantic Segmentation: https://arxiv.org/abs/1907.12273v2

[17] Permutohedral Attention Module for Efficient Non-Local Neural Networks: https://arxiv.org/abs/1907.00641v2

[18] Permutohedral_attention_module: https://github.com/SamuelJoutard/Permutohedral_attention_module

[19] Large Memory Layers with Product Keys: https://arxiv.org/abs/1907.05242v2

[20] XLM: https://github.com/facebookresearch/XLM

[21] Expectation-Maximization Attention Networks for Semantic Segmentation: https://arxiv.org/abs/1907.13426v2

[22] EMANet: https://github.com/XiaLiPKU/EMANet

[23] Compressive Transformers for Long-Range Sequence Modelling: https://arxiv.org/abs/1911.05507v1

[24] compressive-transformer-pytorch: https://github.com/lucidrains/compressive-transformer-pytorch

[25] BP-Transformer: Modelling Long-Range Context via Binary Partitioning: https://arxiv.org/abs/1911.04070v1

[26] BPT: https://github.com/yzh119/BPT

[27] Axial Attention in Multidimensional Transformers: https://arxiv.org/abs/1912.12180v1

[28] axial-attention: https://github.com/lucidrains/axial-attention

[29] Reformer: The Efficient Transformer: https://arxiv.org/abs/2001.04451v2

[30] trax: https://github.com/google/trax/tree/master/trax/models/reformer

[31] Transformer on a Diet: https://arxiv.org/abs/2002.06170v1

[32] transformer-on-diet: https://github.com/cgraywang/transformer-on-diet

[33] Sparse Sinkhorn Attention: https://arxiv.org/abs/2002.11296v1

[34] sinkhorn-transformer: https://github.com/lucidrains/sinkhorn-transformer

[35] SAC: Accelerating and Structuring Self-Attention via Sparse Adaptive Connection: https://arxiv.org/abs/2003.09833v2

[36] Efficient Content-Based Sparse Attention with Routing Transformers: https://arxiv.org/abs/2003.05997v1

[37] routing-transformer: https://github.com/lucidrains/routing-transformer

[38] Longformer: The Long-Document Transformer: https://arxiv.org/abs/2004.05150v1

[39] longformer: https://github.com/allenai/longformer

[40] Neural Architecture Search for Lightweight Non-Local Networks: https://arxiv.org/abs/2004.01961v1

[41] AutoNL: https://github.com/LiYingwei/AutoNL

[42] ETC: Encoding Long and Structured Data in Transformers: https://arxiv.org/abs/2004.08483v2

[43] Multi-scale Transformer Language Models: https://arxiv.org/abs/2005.00581v1

[44] Synthesizer: Rethinking Self-Attention in Transformer Models: https://arxiv.org/abs/2005.00743v1

[45] Jukebox: A Generative Model for Music: https://arxiv.org/abs/2005.00341v1

[46] jukebox: https://github.com/openai/jukebox

[47] GMAT: Global Memory Augmentation for Transformers: https://arxiv.org/abs/2006.03274v1

[48] gmat: https://github.com/ag1988/gmat

[49] Masked Language Modeling for Proteins via Linearly Scalable Long-Context Transformers: https://arxiv.org/abs/2006.03555v1

[50] google-research: https://github.com/google-research/google-research/tree/master/performer/fast_self_attention

[51] Hand-crafted Attention is All You Need? A Study of Attention on Self-supervised Audio Transformer: https://arxiv.org/abs/2006.05174v1

[52] Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention: https://arxiv.org/abs/2006.16236v2

[53] fast-transformers: https://github.com/idiap/fast-transformers

[54] Linformer: Self-Attention with Linear Complexity: https://arxiv.org/abs/2006.04768v3

[55] linformer-pytorch: https://github.com/tatp22/linformer-pytorch

[56] Real-time Semantic Segmentation with Fast Attention: https://arxiv.org/abs/2007.03815v2

[57] Fast Transformers with Clustered Attention: https://arxiv.org/abs/2007.04825v1

[58] fast-transformers: https://github.com/idiap/fast-transformers

[59] Big Bird: Transformers for Longer Sequences: https://arxiv.org/abs/2007.14062v1

[60] A Survey of Long-Term Context in Transformers: https://www.pragmatic.ml/a-survey-of-methods-for-incorporating-long-term-context/

推荐阅读

那些轻轻拍了拍Attention的后浪们
一文看懂 Attention（本质原理+3大优点+5大类型）
消除Aliasing！加州大学&英伟达提出深度学习下采样新思路：自适应低通滤波器层

还有4天！火焰/摔倒/头盔识别等四个赛题即将截止报名

添加极市小助手微信（ID : cvmart2），备注：姓名-学校/公司-研究方向-城市（如：小极-北大-目标检测-深圳），即可申请加入极市目标检测/图像分割/工业检测/人脸/医学影像/3D/SLAM/自动驾驶/超分辨率/姿态估计/ReID/GAN/图像增强/OCR/视频理解等技术交流群：每月大咖直播分享、真实项目需求对接、求职内推、算法竞赛、干货资讯汇总、与 10000+来自港科大、北大、清华、中科院、CMU、腾讯、百度等名校名企视觉开发者互动交流~

△长按添加极市小助手

△长按关注极市平台，获取最新CV干货

觉得有用麻烦给个在看啦~

魔改Attention大集合

Efficient Attention