Large-scale Transformer models bring significant improvements for various downstream vision language tasks with a unified architecture. The performance improvements come with increasing model size, resulting in slow inference speed and increased cost for severing. While some certain predictions benefit from the full complexity of the large-scale model, not all of inputs need the same amount of computation to conduct, potentially leading to computation resource waste. To handle this challenge, early exiting is proposed to adaptively allocate computational power in term of input complexity to improve inference efficiency. The existing early exiting strategies usually adopt output confidence based on intermediate layers as a proxy of input complexity to incur the decision of skipping following layers. However, such strategies cannot apply to encoder in the widely-used unified architecture with both encoder and decoder due to difficulty of output confidence estimation in the encoder. It is suboptimal in term of saving computation power to ignore the early exiting in encoder component. To handle this challenge, we propose a novel early exiting strategy for unified visual language models, which allows dynamically skip the layers in encoder and decoder simultaneously in term of input layer-wise similarities with multiple times of early exiting, namely \textbf{MuE}. By decomposing the image and text modalities in the encoder, MuE is flexible and can skip different layers in term of modalities, advancing the inference efficiency while minimizing performance drop. Experiments on the SNLI-VE and MS COCO datasets show that the proposed approach MuE can reduce expected inference time by up to 50\% and 40\% while maintaining 99\% and 96\% performance respectively.
相關內容
Large language models (LLMs) have demonstrated remarkable capabilities in language generation, understanding, and few-shot learning in recent years. An extensive body of work has explored how their performance may be further improved through the tools of prompting, ranging from verification, self-consistency, or intermediate scratchpads. In this paper, we present a complementary approach to improve language responses where multiple language model instances propose and debate their individual responses and reasoning processes over multiple rounds to arrive at a common final answer. Our findings indicate that this approach significantly enhances mathematical and strategic reasoning across a number of tasks. We also demonstrate that our approach improves the factual validity of generated content, reducing fallacious answers and hallucinations that contemporary models are prone to. Our approach may be directly applied to existing black-box models and uses identical procedure and prompts for all tasks we investigate. Overall, our findings suggest that such "society of minds" approach has the potential to significantly advance the capabilities of LLMs and pave the way for further breakthroughs in language generation and understanding.
Recent advances in state-of-the-art DNN architecture design have been moving toward Transformer models. These models achieve superior accuracy across a wide range of applications. This trend has been consistent over the past several years since Transformer models were originally introduced. However, the amount of compute and bandwidth required for inference of recent Transformer models is growing at a significant rate, and this has made their deployment in latency-sensitive applications challenging. As such, there has been an increased focus on making Transformer models more efficient, with methods that range from changing the architecture design, all the way to developing dedicated domain-specific accelerators. In this work, we survey different approaches for efficient Transformer inference, including: (i) analysis and profiling of the bottlenecks in existing Transformer architectures and their similarities and differences with previous convolutional models; (ii) implications of Transformer architecture on hardware, including the impact of non-linear operations such as Layer Normalization, Softmax, and GELU, as well as linear operations, on hardware design; (iii) approaches for optimizing a fixed Transformer architecture; (iv) challenges in finding the right mapping and scheduling of operations for Transformer models; and (v) approaches for optimizing Transformer models by adapting the architecture using neural architecture search. Finally, we perform a case study by applying the surveyed optimizations on Gemmini, the open-source, full-stack DNN accelerator generator, and we show how each of these approaches can yield improvements, compared to previous benchmark results on Gemmini. Among other things, we find that a full-stack co-design approach with the aforementioned methods can result in up to 88.7x speedup with a minimal performance degradation for Transformer inference.
Although Transformer-based methods have significantly improved state-of-the-art results for long-term series forecasting, they are not only computationally expensive but more importantly, are unable to capture the global view of time series (e.g. overall trend). To address these problems, we propose to combine Transformer with the seasonal-trend decomposition method, in which the decomposition method captures the global profile of time series while Transformers capture more detailed structures. To further enhance the performance of Transformer for long-term prediction, we exploit the fact that most time series tend to have a sparse representation in well-known basis such as Fourier transform, and develop a frequency enhanced Transformer. Besides being more effective, the proposed method, termed as Frequency Enhanced Decomposed Transformer ({\bf FEDformer}), is more efficient than standard Transformer with a linear complexity to the sequence length. Our empirical studies with six benchmark datasets show that compared with state-of-the-art methods, FEDformer can reduce prediction error by $14.8\%$ and $22.6\%$ for multivariate and univariate time series, respectively. the code will be released soon.
With the rise of powerful pre-trained vision-language models like CLIP, it becomes essential to investigate ways to adapt these models to downstream datasets. A recently proposed method named Context Optimization (CoOp) introduces the concept of prompt learning -- a recent trend in NLP -- to the vision domain for adapting pre-trained vision-language models. Specifically, CoOp turns context words in a prompt into a set of learnable vectors and, with only a few labeled images for learning, can achieve huge improvements over intensively-tuned manual prompts. In our study we identify a critical problem of CoOp: the learned context is not generalizable to wider unseen classes within the same dataset, suggesting that CoOp overfits base classes observed during training. To address the problem, we propose Conditional Context Optimization (CoCoOp), which extends CoOp by further learning a lightweight neural network to generate for each image an input-conditional token (vector). Compared to CoOp's static prompts, our dynamic prompts adapt to each instance and are thus less sensitive to class shift. Extensive experiments show that CoCoOp generalizes much better than CoOp to unseen classes, even showing promising transferability beyond a single dataset; and yields stronger domain generalization performance as well. Code is available at //github.com/KaiyangZhou/CoOp.
Transformers have achieved superior performances in many tasks in natural language processing and computer vision, which also intrigues great interests in the time series community. Among multiple advantages of transformers, the ability to capture long-range dependencies and interactions is especially attractive for time series modeling, leading to exciting progress in various time series applications. In this paper, we systematically review transformer schemes for time series modeling by highlighting their strengths as well as limitations through a new taxonomy to summarize existing time series transformers in two perspectives. From the perspective of network modifications, we summarize the adaptations of module level and architecture level of the time series transformers. From the perspective of applications, we categorize time series transformers based on common tasks including forecasting, anomaly detection, and classification. Empirically, we perform robust analysis, model size analysis, and seasonal-trend decomposition analysis to study how Transformers perform in time series. Finally, we discuss and suggest future directions to provide useful research guidance. To the best of our knowledge, this paper is the first work to comprehensively and systematically summarize the recent advances of Transformers for modeling time series data. We hope this survey will ignite further research interests in time series Transformers.
Over the past several years, new machine learning accelerators were being announced and released every month for a variety of applications from speech recognition, video object detection, assisted driving, and many data center applications. This paper updates the survey of AI accelerators and processors from past two years. This paper collects and summarizes the current commercial accelerators that have been publicly announced with peak performance and power consumption numbers. The performance and power values are plotted on a scatter graph, and a number of dimensions and observations from the trends on this plot are again discussed and analyzed. This year, we also compile a list of benchmarking performance results and compute the computational efficiency with respect to peak performance.
Many real-world applications require the prediction of long sequence time-series, such as electricity consumption planning. Long sequence time-series forecasting (LSTF) demands a high prediction capacity of the model, which is the ability to capture precise long-range dependency coupling between output and input efficiently. Recent studies have shown the potential of Transformer to increase the prediction capacity. However, there are several severe issues with Transformer that prevent it from being directly applicable to LSTF, such as quadratic time complexity, high memory usage, and inherent limitation of the encoder-decoder architecture. To address these issues, we design an efficient transformer-based model for LSTF, named Informer, with three distinctive characteristics: (i) a $ProbSparse$ Self-attention mechanism, which achieves $O(L \log L)$ in time complexity and memory usage, and has comparable performance on sequences' dependency alignment. (ii) the self-attention distilling highlights dominating attention by halving cascading layer input, and efficiently handles extreme long input sequences. (iii) the generative style decoder, while conceptually simple, predicts the long time-series sequences at one forward operation rather than a step-by-step way, which drastically improves the inference speed of long-sequence predictions. Extensive experiments on four large-scale datasets demonstrate that Informer significantly outperforms existing methods and provides a new solution to the LSTF problem.
題目： DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference
摘要：
大規模的預訓練語言模型(如BERT)給NLP應用帶來了顯著的改進。然而，它們也因推理速度慢而臭名昭著，這使得它們很難部署到實時應用程序中。我們提出了一個簡單而有效的方法DeeBERT，來加速BERT推理。我們的方法允許樣本在不經過整個模型的情況下提前退出。實驗表明，在模型質量下降最小的情況下，DeeBERT可以節省高達40%的推理時間。進一步的分析顯示了伯特變壓器層的不同行為，并揭示了它們的冗余性。我們的工作為有效地將深度轉換模型應用于下游任務提供了新的思路。
付費5元查看完整內容
Increasing model size when pretraining natural language representations often results in improved performance on downstream tasks. However, at some point further model increases become harder due to GPU/TPU memory limitations, longer training times, and unexpected model degradation. To address these problems, we present two parameter-reduction techniques to lower memory consumption and increase the training speed of BERT. Comprehensive empirical evidence shows that our proposed methods lead to models that scale much better compared to the original BERT. We also use a self-supervised loss that focuses on modeling inter-sentence coherence, and show it consistently helps downstream tasks with multi-sentence inputs. As a result, our best model establishes new state-of-the-art results on the GLUE, RACE, and SQuAD benchmarks while having fewer parameters compared to BERT-large.The code and the pretrained models are available at //github.com/google-research/google-research/tree/master/albert.
Language model pre-training, such as BERT, has significantly improved the performances of many natural language processing tasks. However, pre-trained language models are usually computationally expensive and memory intensive, so it is difficult to effectively execute them on some resource-restricted devices. To accelerate inference and reduce model size while maintaining accuracy, we firstly propose a novel transformer distillation method that is a specially designed knowledge distillation (KD) method for transformer-based models. By leveraging this new KD method, the plenty of knowledge encoded in a large teacher BERT can be well transferred to a small student TinyBERT. Moreover, we introduce a new two-stage learning framework for TinyBERT, which performs transformer distillation at both the pre-training and task-specific learning stages. This framework ensures that TinyBERT can capture both the general-domain and task-specific knowledge of the teacher BERT. TinyBERT is empirically effective and achieves comparable results with BERT in GLUE datasets, while being 7.5x smaller and 9.4x faster on inference. TinyBERT is also significantly better than state-of-the-art baselines, even with only about 28% parameters and 31% inference time of baselines.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191