We propose a routing algorithm that takes a sequence of vectors and computes a new sequence with specified length and vector size. Each output vector maximizes "bang per bit," the difference between a net benefit to use and net cost to ignore data, by better predicting the input vectors. We describe output vectors as geometric objects, as latent variables that assign credit, as query states in a model of associative memory, and as agents in a model of a Society of Mind. We implement the algorithm with optimizations that reduce parameter count, computation, and memory use by orders of magnitude, enabling us to route sequences of greater length than previously possible. We evaluate our implementation on natural language and visual classification tasks, obtaining competitive or state-of-the-art accuracy and end-to-end credit assignments that are interpretable.
translated by 谷歌翻译
Problem instances of a size suitable for practical applications are not likely to be addressed during the noisy intermediate-scale quantum (NISQ) period with (almost) pure quantum algorithms. Hybrid classical-quantum algorithms have potential, however, to achieve good performance on much larger problem instances. We investigate one such hybrid algorithm on a problem of substantial importance: vehicle routing for supply chain logistics with multiple trucks and complex demand structure. We use reinforcement learning with neural networks with embedded quantum circuits. In such neural networks, projecting high-dimensional feature vectors down to smaller vectors is necessary to accommodate restrictions on the number of qubits of NISQ hardware. However, we use a multi-head attention mechanism where, even in classical machine learning, such projections are natural and desirable. We consider data from the truck routing logistics of a company in the automotive sector, and apply our methodology by decomposing into small teams of trucks, and we find results comparable to human truck assignment.
translated by 谷歌翻译
我们提出的方法可以并联有效分类分类。我们的方法将与语义树中给定的节点相对应的一批分类分数和标签转换为与与祖先路径中所有节点相对应的分数和标签硬件加速器。我们在当前的硬件加速器上实现了我们的方法,并用一棵树结合了WordNet 3.0中的所有英语综合体,涵盖了20级的深度,涵盖117,659个类。我们将一批分数和标签转换为各自的祖先路径,从而产生可忽略不计的计算,并且在数据的足迹上仅消耗固定的0.04GB内存。
translated by 谷歌翻译
在深度学习中,模型通常重用所有输入的相同参数。专家的混合(MOE)违反了这一点,而是为每个传入示例选择不同的参数。结果是一个稀疏激活的模型 - 具有残酷数量的参数 - 但恒定的计算成本。然而,尽管MOE取得了一些显着的成功,但复杂性,沟通成本和培训不稳定的阻碍了广泛的采用 - 我们使用Switch Transformer解决了这些领域。我们简化了MOE路由算法和设计直观的改进模型,以降低的通信和计算成本。我们提出的培训技术有助于纠缠不稳定,我们表明稀疏模型可能首次以较低的精度(BFLOAT16)格式进行培训。我们设计了基于T5基数和T5总数的模型,以使用相同的计算资源获得高达7倍的训练速度。这些改进扩展到多语言设置,我们在所有101种语言中衡量对MT5基本版本的收益。最后,我们通过在“巨大的清洁爬行语料库”上预先培训高达数万亿个参数模型,并在T5-XXL模型上实现4倍的速度,从而提高了语言模型的当前规模。
translated by 谷歌翻译
Vehicle routing problems and other combinatorial optimization problems have been approximately solved by reinforcement learning agents with policies based on encoder-decoder models with attention mechanisms. These techniques are of substantial interest but still cannot solve the complex routing problems that arise in a realistic setting which can have many trucks and complex requirements. With the aim of making reinforcement learning a viable technique for supply chain optimization, we develop new extensions to encoder-decoder models for vehicle routing that allow for complex supply chains using classical computing today and quantum computing in the future. We make two major generalizations. First, our model allows for routing problems with multiple trucks. Second, we move away from the simple requirement of having a truck deliver items from nodes to one special depot node, and instead allow for a complex tensor demand structure. We show how our model, even if trained only for a small number of trucks, can be embedded into a large supply chain to yield viable solutions.
translated by 谷歌翻译
多头注意力是最先进的变压器背后的推动力,它在各种自然语言处理(NLP)和计算机视觉任务中实现了出色的性能。已经观察到,对于许多应用,这些注意力头会学习冗余嵌入,并且大多数可以在不降低模型性能的情况下去除。受到这一观察的启发,我们提出了变压器的混合物(变压器-MGK)的混合物,这是一种新型的变压器架构,用每个头部的钥匙混合了变压器中的冗余头部。这些键的混合物遵循高斯混合模型,并使每个注意力头有效地集中在输入序列的不同部分上。与传统的变压器对应物相比,变压器-MGK会加速训练和推理,具有较少的参数,并且需要更少的拖船来计算,同时实现跨任务的可比性或更高的准确性。 Transformer-MGK也可以轻松扩展到线性注意力。我们从经验上证明了在一系列实用应用中变形金属MGK的优势,包括语言建模和涉及非常长序列的任务。在Wikitext-103和远程竞技场基准中,具有4个头部的变压器MGK具有与基线变压器具有8个头的可比性或更好的性能。
translated by 谷歌翻译
在本文中,我们试图通过引入深度学习模型的句法归纳偏见来建立两所学校之间的联系。我们提出了两个归纳偏见的家族,一个家庭用于选区结构,另一个用于依赖性结构。选区归纳偏见鼓励深度学习模型使用不同的单位(或神经元)分别处理长期和短期信息。这种分离为深度学习模型提供了一种方法,可以从顺序输入中构建潜在的层次表示形式,即更高级别的表示由高级表示形式组成,并且可以分解为一系列低级表示。例如,在不了解地面实际结构的情况下,我们提出的模型学会通过根据其句法结构组成变量和运算符的表示来处理逻辑表达。另一方面,依赖归纳偏置鼓励模型在输入序列中找到实体之间的潜在关系。对于自然语言,潜在关系通常被建模为一个定向依赖图,其中一个单词恰好具有一个父节点和零或几个孩子的节点。将此约束应用于类似变压器的模型之后,我们发现该模型能够诱导接近人类专家注释的有向图,并且在不同任务上也优于标准变压器模型。我们认为,这些实验结果为深度学习模型的未来发展展示了一个有趣的选择。
translated by 谷歌翻译
深度学习使用由其重量进行参数化的神经网络。通常通过调谐重量来直接最小化给定损耗功能来训练神经网络。在本文中,我们建议将权重重新参数转化为网络中各个节点的触发强度的目标。给定一组目标,可以计算使得发射强度最佳地满足这些目标的权重。有人认为,通过我们称之为级联解压缩的过程,使用培训的目标解决爆炸梯度的问题,并使损失功能表面更加光滑,因此导致更容易,培训更快,以及潜在的概括,神经网络。它还允许更容易地学习更深层次和经常性的网络结构。目标对重量的必要转换有额外的计算费用,这是在许多情况下可管理的。在目标空间中学习可以与现有的神经网络优化器相结合,以额外收益。实验结果表明了使用目标空间的速度,以及改进的泛化的示例,用于全连接的网络和卷积网络,以及调用和处理长时间序列的能力,并使用经常性网络进行自然语言处理。
translated by 谷歌翻译
神经网络是强大的功能估计器,导致其作为建模结构化数据的首选范式的地位。但是,与其他强调问题模块化的结构化表示不同,例如因子图 - 神经网络通常是从输入到输出的单片映射,并具有固定的计算顺序。这种限制阻止他们捕获模型变量之间的不同计算方向和相互作用。在本文中,我们结合了因子图和神经网络的代表性强度,提出了无向神经网络(UNNS):一个灵活的框架,用于指定可以按任何顺序执行的计算。对于特定的选择,我们提出的模型集成并扩展了许多现有的架构:带有隐式层的馈电,经常性,自我发项网络,自动编码器和网络。我们在一系列任务上展示了无方向性的神经体系结构的有效性:受树约束依赖性解析,卷积图像分类和序列完成。通过改变计算顺序,我们展示了如何同时将单个UNN用作分类器和原型发生器,以及它如何填充输入序列的缺失部分,从而使它们成为进一步研究的有希望的领域。
translated by 谷歌翻译
我用Hunglish2语料库训练神经电脑翻译任务的模型。这项工作的主要贡献在培训NMT模型期间评估不同的数据增强方法。我提出了5种不同的增强方法,这些方法是结构感知的,这意味着而不是随机选择用于消隐或替换的单词,句子的依赖树用作增强的基础。我首先关于神经网络的详细文献综述,顺序建模,神经机翻译,依赖解析和数据增强。经过详细的探索性数据分析和Hunglish2语料库的预处理之后,我使用所提出的数据增强技术进行实验。匈牙利语的最佳型号达到了33.9的BLEU得分,而英国匈牙利最好的模型达到了28.6的BLEU得分。
translated by 谷歌翻译
稀疏的专家模型是一个三十年来的概念,作为深度学习中流行的建筑。这类体系结构包括专家的混合物,交换变压器,路由网络,基础层等,所有这些都以一个统一的想法,即每个示例都由参数的一个子集进行。通过这样做,稀疏度将参数计数与每个示例的计算分解,从而允许使用极大但有效的模型。最终的模型显示了各种领域的显着改善,例如自然语言处理,计算机视觉和语音识别。我们回顾了稀疏专家模型的概念,提供了对常见算法的基本描述,将深度学习时代的进步进行上下文化,并通过突出未来工作的领域来结束。
translated by 谷歌翻译
我们介绍了一个非常简单的注意力,以便对序列长度的级别长度和延伸,以便对需要$ o(\ log n)$内存的自我关注的扩展来表示。与经常说的信念相比,自我注意需要$ O(n ^ 2)$内存。虽然时间复杂性仍然是$ O(n ^ 2)$,但是设备存储器而不是计算能力通常是现代加速器上的限制因素。因此,减少注意力的记忆要求允许处理比否则可以是可行的更长的序列。我们为需要$ O(\ sqrt {n})$内存的加速器提供实际实施,是数字稳定的,并且在标准实施的运行时间的几乎没有百分比范围内。我们还演示了如何区分功能,同时剩余内存效率。对于序列长度为16384,自我注意的存储器开销减少59倍,用于推断和32倍以进行分化。
translated by 谷歌翻译
Transformers-based models, such as BERT, have been one of the most successful deep learning models for NLP. Unfortunately, one of their core limitations is the quadratic dependency (mainly in terms of memory) on the sequence length due to their full attention mechanism. To remedy this, we propose, BIGBIRD, a sparse attention mechanism that reduces this quadratic dependency to linear. We show that BIGBIRD is a universal approximator of sequence functions and is Turing complete, thereby preserving these properties of the quadratic, full attention model. Along the way, our theoretical analysis reveals some of the benefits of having O(1) global tokens (such as CLS), that attend to the entire sequence as part of the sparse attention mechanism. The proposed sparse attention can handle sequences of length up to 8x of what was previously possible using similar hardware. As a consequence of the capability to handle longer context, BIGBIRD drastically improves performance on various NLP tasks such as question answering and summarization. We also propose novel applications to genomics data.
translated by 谷歌翻译
Self-attention has recently been adopted for a wide range of sequence modeling problems. Despite its effectiveness, self-attention suffers from quadratic compute and memory requirements with respect to sequence length. Successful approaches to reduce this complexity focused on attending to local sliding windows or a small set of locations independent of content. Our work proposes to learn dynamic sparse attention patterns that avoid allocating computation and memory to attend to content unrelated to the query of interest. This work builds upon two lines of research: it combines the modeling flexibility of prior work on content-based sparse attention with the efficiency gains from approaches based on local, temporal sparse attention. Our model, the Routing Transformer, endows selfattention with a sparse routing module based on online k-means while reducing the overall complexity of attention to O(n 1.5 d) from O(n 2 d) for sequence length n and hidden dimension d. We show that our model outperforms comparable sparse attention models on language modeling on Wikitext-103 (15.8 vs 18.3 perplexity), as well as on image generation on ImageNet-64 (3.43 vs 3.44 bits/dim) while using fewer self-attention layers. Additionally, we set a new state-of-the-art on the newly released PG-19 data-set, obtaining a test perplexity of 33.2 with a 22 layer Routing Transformer model trained on sequences of length 8192. We open-source the code for Routing Transformer in Tensorflow. *
translated by 谷歌翻译
在这项工作中,我们介绍了内核化变压器,这是一个通用,可扩展的,数据驱动的框架,用于学习变压器中的内核功能。我们的框架将变压器内核作为光谱特征图之间的点产物近似,并通过学习光谱分布来学习内核。这不仅有助于学习通用的内核端到端,而且还可以减少变压器从二次到线性的时间和空间复杂性。我们表明,在准确性和计算效率方面,内核化的变压器实现了与现有的有效变压器体系结构相当的性能。我们的研究还表明,内核的选择对性能有重大影响,而内核学习变体是固定内核变压器的竞争替代方案,无论是长时间的序列任务。
translated by 谷歌翻译
我们介绍了块状变压器,该变压器以序列的反复方式应用变压器层,并且相对于序列长度具有线性复杂性。我们的复发单元在训练过程中在代币的块而不是单个令牌上运行,并利用块内并行计算,以便有效利用加速器硬件。单元本身非常简单。它仅仅是一个变压器层:它使用自我注意事项和交叉注意力来有效计算大量状态向量和令牌上的复发函数。我们的设计部分受到LSTM单元的启发,它使用LSTM风格的大门,但它可以将典型的LSTM单元缩放为几个数量级。我们的复发实现在计算时间和参数计数中都具有相同的成本作为传统的变压器层,但是在很长的序列中,语言建模任务中的语言建模任务的困惑极大地改善了。我们的模型比远程变压器XL基线的表现宽大,同时运行的速度是两倍。我们证明了它在PG19(书籍),Arxiv论文和GitHub源代码上的有效性。我们的代码已发布为开​​源。
translated by 谷歌翻译
变压器架构现在是序列建模任务的核心。注意机制是核心,它可以在序列中对长期依赖性进行有效的建模。最近,变压器已成功地应用于计算机视觉域,在该域中首先将2D图像分割成斑块,然后将其视为1D序列。然而,这种线性化会损害图像中空间位置的概念,该图像具有重要的视觉线索。为了弥合差距,我们提出了连锁反应,这是视觉变压器的次级注意机制。基于最近基于内核的有效注意机制,我们设计了一种新型的动态编程算法,该算法将不同令牌的贡献加重了与它们在线性观察到的2D空间中相对空间距离的查询的贡献。广泛的实验和分析证明了连锁反应对各种视觉任务的有效性。
translated by 谷歌翻译
We present a new approach to modeling sequential data: the deep equilibrium model (DEQ). Motivated by an observation that the hidden layers of many existing deep sequence models converge towards some fixed point, we propose the DEQ approach that directly finds these equilibrium points via root-finding. Such a method is equivalent to running an infinite depth (weight-tied) feedforward network, but has the notable advantage that we can analytically backpropagate through the equilibrium point using implicit differentiation. Using this approach, training and prediction in these networks require only constant memory, regardless of the effective "depth" of the network. We demonstrate how DEQs can be applied to two state-of-the-art deep sequence models: self-attention transformers and trellis networks. On large-scale language modeling tasks, such as the WikiText-103 benchmark, we show that DEQs 1) often improve performance over these stateof-the-art models (for similar parameter counts); 2) have similar computational requirements to existing models; and 3) vastly reduce memory consumption (often the bottleneck for training large sequence models), demonstrating an up-to 88% memory reduction in our experiments. The code is available at https://github. com/locuslab/deq.
translated by 谷歌翻译
Large Transformer models routinely achieve state-of-the-art results on a number of tasks but training these models can be prohibitively costly, especially on long sequences. We introduce two techniques to improve the efficiency of Transformers. For one, we replace dot-product attention by one that uses locality-sensitive hashing, changing its complexity from O(L 2 ) to O(L log L), where L is the length of the sequence. Furthermore, we use reversible residual layers instead of the standard residuals, which allows storing activations only once in the training process instead of N times, where N is the number of layers. The resulting model, the Reformer, performs on par with Transformer models while being much more memory-efficient and much faster on long sequences.
translated by 谷歌翻译
这项正在进行的工作旨在为统计学习提供统一的介绍,从诸如GMM和HMM等经典模型到现代神经网络(如VAE和扩散模型)缓慢地构建。如今,有许多互联网资源可以孤立地解释这一点或新的机器学习算法,但是它们并没有(也不能在如此简短的空间中)将这些算法彼此连接起来,或者与统计模型的经典文献相连现代算法出现了。同样明显缺乏的是一个单一的符号系统,尽管对那些已经熟悉材料的人(如这些帖子的作者)不满意,但对新手的入境造成了重大障碍。同样,我的目的是将各种模型(尽可能)吸收到一个用于推理和学习的框架上,表明(以及为什么)如何以最小的变化将一个模型更改为另一个模型(其中一些是新颖的,另一些是文献中的)。某些背景当然是必要的。我以为读者熟悉基本的多变量计算,概率和统计以及线性代数。这本书的目标当然不是​​完整性,而是从基本知识到过去十年中极强大的新模型的直线路径或多或少。然后,目标是补充而不是替换,诸如Bishop的\ emph {模式识别和机器学习}之类的综合文本,该文本现在已经15岁了。
translated by 谷歌翻译