Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userMemory-Efficient Differentiable Transformer Architecture Search
349
0
0.0
(
0
)
Added by
Yuekai Zhao
Publication date
2021
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Differentiable architecture search (DARTS) is successfully applied in many vision tasks. However, directly using DARTS for Transformers is memory-intensive, which renders the search process infeasible. To this end, we propose a multi-split reversible network and combine it with DARTS. Specifically, we devise a backpropagation-with-reconstruction algorithm so that we only need to store the last layers outputs. By relieving the memory burden for DARTS, it allows us to search with larger hidden size and more candidate operations. We evaluate the searched architecture on three sequence-to-sequence datasets, i.e., WMT14 English-German, WMT14 English-French, and WMT14 English-Czech. Experimental results show that our network consistently outperforms standard Transformers across the tasks. Moreover, our method compares favorably with big-size Evolved Transformers, reducing search computation by an order of magnitude.
rate research
Read More
Differentiable neural architecture search (DNAS) is known for its capacity in the automatic generation of superior neural networks. However, DNAS based methods suffer from memory usage explosion when the search space expands, which may prevent them from running successfully on even advanced GPU platforms. On the other hand, reinforcement learning (RL) based methods, while being memory efficient, are extremely time-consuming. Combining the advantages of both types of methods, this paper presents RADARS, a scalable RL-aided DNAS framework that can explore large search spaces in a fast and memory-efficient manner. RADARS iteratively applies RL to prune undesired architecture candidates and identifies a promising subspace to carry out DNAS. Experiments using a workstation with 12 GB GPU memory show that on CIFAR-10 and ImageNet datasets, RADARS can achieve up to 3.41% higher accuracy with 2.5X search time reduction compared with a state-of-the-art RL-based method, while the two DNAS baselines cannot complete due to excessive memory usage or search time. To the best of the authors knowledge, this is the first DNAS framework that can handle large search spaces with bounded memory usage.
Neural architecture search (NAS) aims to discover network architectures with desired properties such as high accuracy or low latency. Recently, differentiable NAS (DNAS) has demonstrated promising results while maintaining a search cost orders of magnitude lower than reinforcement learning (RL) based NAS. However, DNAS models can only optimize differentiable loss functions in search, and they require an accurate differentiable approximation of non-differentiable criteria. In this work, we present UNAS, a unified framework for NAS, that encapsulates recent DNAS and RL-based approaches under one framework. Our framework brings the best of both worlds, and it enables us to search for architectures with both differentiable and non-differentiable criteria in one unified framework while maintaining a low search cost. Further, we introduce a new objective function for search based on the generalization gap that prevents the selection of architectures prone to overfitting. We present extensive experiments on the CIFAR-10, CIFAR-100, and ImageNet datasets and we perform search in two fundamentally different search spaces. We show that UNAS obtains the state-of-the-art average accuracy on all three datasets when compared to the architectures searched in the DARTS space. Moreover, we show that UNAS can find an efficient and accurate architecture in the ProxylessNAS search space, that outperforms existing MobileNetV2 based architectures. The source code is available at https://github.com/NVlabs/unas .
The Transformer is widely used in natural language processing tasks. To train a Transformer however, one usually needs a carefully designed learning rate warm-up stage, which is shown to be crucial to the final performance but will slow down the optimization and bring more hyper-parameter tunings. In this paper, we first study theoretically why the learning rate warm-up stage is essential and show that the location of layer normalization matters. Specifically, we prove with mean field theory that at initialization, for the original-designed Post-LN Transformer, which places the layer normalization between the residual blocks, the expected gradients of the parameters near the output layer are large. Therefore, using a large learning rate on those gradients makes the training unstable. The warm-up stage is practically helpful for avoiding this problem. On the other hand, our theory also shows that if the layer normalization is put inside the residual blocks (recently proposed as Pre-LN Transformer), the gradients are well-behaved at initialization. This motivates us to remove the warm-up stage for the training of Pre-LN Transformers. We show in our experiments that Pre-LN Transformers without the warm-up stage can reach comparable results with baselines while requiring significantly less training time and hyper-parameter tuning on a wide range of applications.
textit{Differentiable ARchiTecture Search} (DARTS) has recently become the mainstream of neural architecture search (NAS) due to its efficiency and simplicity. With a gradient-based bi-level optimization, DARTS alternately optimizes the inner model weights and the outer architecture parameter in a weight-sharing supernet. A key challenge to the scalability and quality of the learned architectures is the need for differentiating through the inner-loop optimisation. While much has been discussed about several potentially fatal factors in DARTS, the architecture gradient, a.k.a. hypergradient, has received less attention. In this paper, we tackle the hypergradient computation in DARTS based on the implicit function theorem, making it only depends on the obtained solution to the inner-loop optimization and agnostic to the optimization path. To further reduce the computational requirements, we formulate a stochastic hypergradient approximation for differentiable NAS, and theoretically show that the architecture optimization with the proposed method, named iDARTS, is expected to converge to a stationary point. Comprehensive experiments on two NAS benchmark search spaces and the common NAS search space verify the effectiveness of our proposed method. It leads to architectures outperforming, with large margins, those learned by the baseline methods.
Differentiable Architecture Search (DARTS) is a recently proposed neural architecture search (NAS) method based on a differentiable relaxation. Due to its success, numerous variants analyzing and improving parts of the DARTS framework have recently been proposed. By considering the problem as a constrained bilevel optimization, we propose and analyze three improvements to architectural weight competition, update scheduling, and regularization towards discretization. First, we introduce a new approach to the activation of architecture weights, which prevents confounding competition within an edge and allows for fair comparison across edges to aid in discretization. Next, we propose a dynamic schedule based on per-minibatch network information to make architecture updates more informed. Finally, we consider two regularizations, based on proximity to discretization and the Alternating Directions Method of Multipliers (ADMM) algorithm, to promote early discretization. Our results show that this new activation scheme reduces final architecture size and the regularizations improve reliability in search results while maintaining comparable performance to state-of-the-art in NAS, especially when used with our new dynamic informed schedule.
suggested questions
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
