Search

A visually rich document (VRD) utilizes visual features along with linguistic cues to disseminate information. Training a custom extractor that identifies named entities from a document requires a large number of instances of the target document type annotated at textual and visual modalities. This is an expensive bottleneck in enterprise scenarios, where we want to train custom extractors for thousands of different document types in a scalable way. Pre-training an extractor model on unlabeled instances of the target document type, followed by a fine-tuning step on human-labeled instances does not work in these scenarios, as it surpasses the maximum allowable training time allocated for the extractor. We address this scenario by proposing a Noise-Aware Training method or NAT in this paper. Instead of acquiring expensive human-labeled documents, NAT utilizes weakly labeled documents to train an extractor in a scalable way. To avoid degradation in the model's quality due to noisy, weakly labeled samples, NAT estimates the confidence of each training sample and incorporates it as uncertainty measure during training. We train multiple state-of-the-art extractor models using NAT. Experiments on a number of publicly available and in-house datasets show that NAT-trained models are not only robust in performance -- it outperforms a transfer-learning baseline by up to 6% in terms of macro-F1 score, but it is also more label-efficient -- it reduces the amount of human-effort required to obtain comparable performance by up to 73%.

We present Gecko, a compact and versatile text embedding model. Gecko achieves strong retrieval performance by leveraging a key idea: distilling knowledge from large language models (LLMs) into a retriever. Our two-step distillation process begins with generating diverse, synthetic paired data using an LLM. Next, we further refine the data quality by retrieving a set of candidate passages for each query, and relabeling the positive and hard negative passages using the same LLM. The effectiveness of our approach is demonstrated by the compactness of the Gecko. On the Massive Text Embedding Benchmark (MTEB), Gecko with 256 embedding dimensions outperforms all existing entries with 768 embedding size. Gecko with 768 embedding dimensions achieves an average score of 66.31, competing with 7x larger models and 5x higher dimensional embeddings., Comment: 18 pages

Continual learning aims to enable a single model to learn a sequence of tasks without catastrophic forgetting. Top-performing methods usually require a rehearsal buffer to store past pristine examples for experience replay, which, however, limits their practical value due to privacy and memory constraints. In this work, we present a simple yet effective framework, DualPrompt, which learns a tiny set of parameters, called prompts, to properly instruct a pre-trained model to learn tasks arriving sequentially without buffering past examples. DualPrompt presents a novel approach to attach complementary prompts to the pre-trained backbone, and then formulates the objective as learning task-invariant and task-specific "instructions". With extensive experimental validation, DualPrompt consistently sets state-of-the-art performance under the challenging class-incremental setting. In particular, DualPrompt outperforms recent advanced continual learning methods with relatively large buffer sizes. We also introduce a more challenging benchmark, Split ImageNet-R, to help generalize rehearsal-free continual learning research. Source code is available at https://github.com/google-research/l2p., Comment: Published at ECCV 2022 as a conference paper

The mainstream paradigm behind continual learning has been to adapt the model parameters to non-stationary data distributions, where catastrophic forgetting is the central challenge. Typical methods rely on a rehearsal buffer or known task identity at test time to retrieve learned knowledge and address forgetting, while this work presents a new paradigm for continual learning that aims to train a more succinct memory system without accessing task identity at test time. Our method learns to dynamically prompt (L2P) a pre-trained model to learn tasks sequentially under different task transitions. In our proposed framework, prompts are small learnable parameters, which are maintained in a memory space. The objective is to optimize prompts to instruct the model prediction and explicitly manage task-invariant and task-specific knowledge while maintaining model plasticity. We conduct comprehensive experiments under popular image classification benchmarks with different challenging continual learning settings, where L2P consistently outperforms prior state-of-the-art methods. Surprisingly, L2P achieves competitive results against rehearsal-based methods even without a rehearsal buffer and is directly applicable to challenging task-agnostic continual learning. Source code is available at https://github.com/google-research/l2p., Comment: Published at CVPR 2022 as a conference paper

Scheduling precedence-constrained tasks is a classical problem that has been studied for more than fifty years. However, little progress has been made in the setting where there are communication delays between tasks. Results for the case of identical machines were derived nearly thirty years ago, and yet no results for related machines have followed. In this work, we propose a new scheduler, Generalized Earliest Time First (GETF), and provide the first provable, worst-case approximation guarantees for the goals of minimizing both the makespan and total weighted completion time of tasks with precedence constraints on related machines with machine-dependent communication times.

This paper studies two design tasks faced by a geo-distributed cloud data market: which data to purchase (data purchasing) and where to place/replicate the data for delivery (data placement). We show that the joint problem of data purchasing and data placement within a cloud data market can be viewed as a facility location problem, and is thus NP-hard. However, we give a provably optimal algorithm for the case of a data market made up of a single data center, and then generalize the structure from the single data center setting in order to develop a near-optimal, polynomial-time algorithm for a geo-distributed data market. The resulting design, Datum, decomposes the joint purchasing and placement problem into two subproblems, one for data purchasing and one for data placement, using a transformation of the underlying bandwidth costs. We show, via a case study, that Datum is near-optimal (within 1.6%) in practical settings., Comment: 35 pages, 4 figures

Data centers have emerged as promising resources for demand response, particularly for emergency demand response (EDR), which saves the power grid from incurring blackouts during emergency situations. However, currently, data centers typically participate in EDR by turning on backup (diesel) generators, which is both expensive and environmentally unfriendly. In this paper, we focus on "greening" demand response in multi-tenant data centers, i.e., colocation data centers, by designing a pricing mechanism through which the data center operator can efficiently extract load reductions from tenants during emergency periods to fulfill energy reduction requirement for EDR. In particular, we propose a pricing mechanism for both mandatory and voluntary EDR programs, ColoEDR, that is based on parameterized supply function bidding and provides provably near-optimal efficiency guarantees, both when tenants are price-taking and when they are price-anticipating. In addition to analytic results, we extend the literature on supply function mechanism design, and evaluate ColoEDR using trace-based simulation studies. These validate the efficiency analysis and conclude that the pricing mechanism is both beneficial to the environment and to the data center operator (by decreasing the need for backup diesel generation), while also aiding tenants (by providing payments for load reductions)., Comment: 34 pages, 6 figures