Revisiting SLO and Goodput Metrics in LLM Serving

Large language models (LLMs) have achieved remarkable performance and are widely deployed in various applications, while the serving of LLM inference has raised concerns about user experience and serving throughput. Accordingly, service level objectives (SLOs) and goodput-the number of requests that meet SLOs per second-are introduced to evaluate the performance of LLM serving. However, existing metrics fail to capture the nature of user experience. We observe two ridiculous phenomena in existing metrics: 1) delaying token delivery can smooth the tail time between tokens (tail TBT) of a request and 2) dropping the request that fails to meet the SLOs midway can improve goodput. In this paper, we revisit SLO and goodput metrics in LLM serving and propose a unified metric framework smooth goodput including SLOs and goodput to reflect the nature of user experience in LLM serving. The framework can adapt to specific goals of different tasks by setting parameters. We re-evaluate the performance of different LLM serving systems under multiple workloads based on this unified framework and provide possible directions for future optimization of existing strategies. We hope that this framework can provide a unified standard for evaluating LLM serving and foster researches in the field of LLM serving optimization to move in a cohesive direction.

MedImageInsight: An Open-Source Embedding Model for General Domain Medical Imaging

In this work, we present MedImageInsight, an open-source medical imaging embedding model. MedImageInsight is trained on medical images with associated text and labels across a diverse collection of domains, including X-Ray, CT, MRI, dermoscopy, OCT, fundus photography, ultrasound, histopathology, and mammography. Rigorous evaluations demonstrate MedImageInsight's ability to achieve state-of-the-art (SOTA) or human expert level performance across classification, image-image search, and fine-tuning tasks. Specifically, on public datasets, MedImageInsight achieves SOTA in CT 3D medical image retrieval, as well as SOTA in disease classification and search for chest X-ray, dermatology, and OCT imaging. Furthermore, MedImageInsight achieves human expert performance in bone age estimation (on both public and partner data), as well as AUC above 0.9 in most other domains. When paired with a text decoder, MedImageInsight achieves near SOTA level single image report findings generation with less than 10\% the parameters of other models. Compared to fine-tuning GPT-4o with only MIMIC-CXR data for the same task, MedImageInsight outperforms in clinical metrics, but underperforms on lexical metrics where GPT-4o sets a new SOTA. Importantly for regulatory purposes, MedImageInsight can generate ROC curves, adjust sensitivity and specificity based on clinical need, and provide evidence-based decision support through image-image search (which can also enable retrieval augmented generation). In an independent clinical evaluation of image-image search in chest X-ray, MedImageInsight outperformed every other publicly available foundation model evaluated by large margins (over 6 points AUC), and significantly outperformed other models in terms of AI fairness (across age and gender). We hope releasing MedImageInsight will help enhance collective progress in medical imaging AI research and development.

TokenBinder: Text-Video Retrieval with One-to-Many Alignment Paradigm

Text-Video Retrieval (TVR) methods typically match query-candidate pairs by aligning text and video features in coarse-grained, fine-grained, or combined (coarse-to-fine) manners. However, these frameworks predominantly employ a one(query)-to-one(candidate) alignment paradigm, which struggles to discern nuanced differences among candidates, leading to frequent mismatches. Inspired by Comparative Judgement in human cognitive science, where decisions are made by directly comparing items rather than evaluating them independently, we propose TokenBinder. This innovative two-stage TVR framework introduces a novel one-to-many coarse-to-fine alignment paradigm, imitating the human cognitive process of identifying specific items within a large collection. Our method employs a Focused-view Fusion Network with a sophisticated cross-attention mechanism, dynamically aligning and comparing features across multiple videos to capture finer nuances and contextual variations. Extensive experiments on six benchmark datasets confirm that TokenBinder substantially outperforms existing state-of-the-art methods. These results demonstrate its robustness and the effectiveness of its fine-grained alignment in bridging intra- and inter-modality information gaps in TVR tasks.

Contrastive Federated Learning with Tabular Data Silos

Learning from data silos is a difficult task for organizations that need to obtain knowledge of objects that appeared in multiple independent data silos. Objects in multi-organizations, such as government agents, are referred by different identifiers, such as driver license, passport number, and tax file number. The data distributions in data silos are mostly non-IID (Independently and Identically Distributed), labelless, and vertically partitioned (i.e., having different attributes). Privacy concerns harden the above issues. Conditions inhibit enthusiasm for collaborative work. While Federated Learning (FL) has been proposed to address these issues, the difficulty of labeling, namely, label costliness, often hinders optimal model performance. A potential solution lies in contrastive learning, an unsupervised self-learning technique to represent semantic data by contrasting similar data pairs. However, contrastive learning is currently not designed to handle tabular data silos that existed within multiple organizations where data linkage by quasi identifiers are needed. To address these challenges, we propose using semi-supervised contrastive federated learning, which we refer to as Contrastive Federated Learning with Data Silos (CFL). Our approach tackles the aforementioned issues with an integrated solution. Our experimental results demonstrate that CFL outperforms current methods in addressing these challenges and providing improvements in accuracy. Additionally, we present positive results that showcase the advantages of our contrastive federated learning approach in complex client environments.

Optimizing Novelty of Top-k Recommendations using Large Language Models and Reinforcement Learning

Given an input query, a recommendation model is trained using user feedback data (e.g., click data) to output a ranked list of items. In real-world systems, besides accuracy, an important consideration for a new model is novelty of its top-k recommendations w.r.t. an existing deployed model. However, novelty of top-k items is a difficult goal to optimize a model for, since it involves a non-differentiable sorting operation on the model's predictions. Moreover, novel items, by definition, do not have any user feedback data. Given the semantic capabilities of large language models, we address these problems using a reinforcement learning (RL) formulation where large language models provide feedback for the novel items. However, given millions of candidate items, the sample complexity of a standard RL algorithm can be prohibitively high. To reduce sample complexity, we reduce the top-k list reward to a set of item-wise rewards and reformulate the state space to consist of <query, item> tuples such that the action space is reduced to a binary decision; and show that this reformulation results in a significantly lower complexity when the number of items is large. We evaluate the proposed algorithm on improving novelty for a query-ad recommendation task on a large-scale search engine. Compared to supervised finetuning on recent <query, ad> pairs, the proposed RL-based algorithm leads to significant novelty gains with minimal loss in recall. We obtain similar results on the ORCAS query-webpage matching dataset and a product recommendation dataset based on Amazon reviews.

CaseGNN++: Graph Contrastive Learning for Legal Case Retrieval with Graph Augmentation

Legal case retrieval (LCR) is a specialised information retrieval task that aims to find relevant cases to a given query case. LCR holds pivotal significance in facilitating legal practitioners in finding precedents. Most of existing LCR methods are based on traditional lexical models and language models, which have gained promising performance in retrieval. However, the domain-specific structural information inherent in legal documents is yet to be exploited to further improve the performance. Our previous work CaseGNN successfully harnesses text-attributed graphs and graph neural networks to address the problem of legal structural information neglect. Nonetheless, there remain two aspects for further investigation: (1) The underutilization of rich edge information within text-attributed case graphs limits CaseGNN to generate informative case representation. (2) The inadequacy of labelled data in legal datasets hinders the training of CaseGNN model. In this paper, CaseGNN++, which is extended from CaseGNN, is proposed to simultaneously leverage the edge information and additional label data to discover the latent potential of LCR models. Specifically, an edge feature-based graph attention layer (EUGAT) is proposed to comprehensively update node and edge features during graph modelling, resulting in a full utilisation of structural information of legal cases. Moreover, a novel graph contrastive learning objective with graph augmentation is developed in CaseGNN++ to provide additional training signals, thereby enhancing the legal comprehension capabilities of CaseGNN++ model. Extensive experiments on two benchmark datasets from COLIEE 2022 and COLIEE 2023 demonstrate that CaseGNN++ not only significantly improves CaseGNN but also achieves supreme performance compared to state-of-the-art LCR methods. Code has been released on https://github.com/yanran-tang/CaseGNN.

On-device Online Learning and Semantic Management of TinyML Systems

Recent advances in Tiny Machine Learning (TinyML) empower low-footprint embedded devices for real-time on-device Machine Learning. While many acknowledge the potential benefits of TinyML, its practical implementation presents unique challenges. This study aims to bridge the gap between prototyping single TinyML models and developing reliable TinyML systems in production: (1) Embedded devices operate in dynamically changing conditions. Existing TinyML solutions primarily focus on inference, with models trained offline on powerful machines and deployed as static objects. However, static models may underperform in the real world due to evolving input data distributions. We propose online learning to enable training on constrained devices, adapting local models towards the latest field conditions. (2) Nevertheless, current on-device learning methods struggle with heterogeneous deployment conditions and the scarcity of labeled data when applied across numerous devices. We introduce federated meta-learning incorporating online learning to enhance model generalization, facilitating rapid learning. This approach ensures optimal performance among distributed devices by knowledge sharing. (3) Moreover, TinyML's pivotal advantage is widespread adoption. Embedded devices and TinyML models prioritize extreme efficiency, leading to diverse characteristics ranging from memory and sensors to model architectures. Given their diversity and non-standardized representations, managing these resources becomes challenging as TinyML systems scale up. We present semantic management for the joint management of models and devices at scale. We demonstrate our methods through a basic regression example and then assess them in three real-world TinyML applications: handwritten character image classification, keyword audio classification, and smart building presence detection, confirming our approaches' effectiveness.

High Order Reasoning for Time Critical Recommendation in Evidence-based Medicine

In time-critical decisions, human decision-makers can interact with AI-enabled situation-aware software to evaluate many imminent and possible scenarios, retrieve billions of facts, and estimate different outcomes based on trillions of parameters in a fraction of a second. In high-order reasoning, "what-if" questions can be used to challenge the assumptions or pre-conditions of the reasoning, "why-not" questions can be used to challenge on the method applied in the reasoning, "so-what" questions can be used to challenge the purpose of the decision, and "how-about" questions can be used to challenge the applicability of the method. When above high-order reasoning questions are applied to assist human decision-making, it can help humans to make time-critical decisions and avoid false-negative or false-positive types of errors. In this paper, we present a model of high-order reasoning to offer recommendations in evidence-based medicine in a time-critical fashion for the applications in ICU. The Large Language Model (LLM) is used in our system. The experiments demonstrated the LLM exhibited optimal performance in the "What-if" scenario, achieving a similarity of 88.52% with the treatment plans of human doctors. In the "Why-not" scenario, the best-performing model tended to opt for alternative treatment plans in 70% of cases for patients who died after being discharged from the ICU. In the "So-what" scenario, the optimal model provided a detailed analysis of the motivation and significance of treatment plans for ICU patients, with its reasoning achieving a similarity of 55.6% with actual diagnostic information. In the "How-about" scenario, the top-performing LLM demonstrated a content similarity of 66.5% in designing treatment plans transferring for similar diseases. Meanwhile, LLMs managed to predict the life status of patients after their discharge from the ICU with an accuracy of 70%.

CaseLink: Inductive Graph Learning for Legal Case Retrieval

In case law, the precedents are the relevant cases that are used to support the decisions made by the judges and the opinions of lawyers towards a given case. This relevance is referred to as the case-to-case reference relation. To efficiently find relevant cases from a large case pool, retrieval tools are widely used by legal practitioners. Existing legal case retrieval models mainly work by comparing the text representations of individual cases. Although they obtain a decent retrieval accuracy, the intrinsic case connectivity relationships among cases have not been well exploited for case encoding, therefore limiting the further improvement of retrieval performance. In a case pool, there are three types of case connectivity relationships: the case reference relationship, the case semantic relationship, and the case legal charge relationship. Due to the inductive manner in the task of legal case retrieval, using case reference as input is not applicable for testing. Thus, in this paper, a CaseLink model based on inductive graph learning is proposed to utilise the intrinsic case connectivity for legal case retrieval, a novel Global Case Graph is incorporated to represent both the case semantic relationship and the case legal charge relationship. A novel contrastive objective with a regularisation on the degree of case nodes is proposed to leverage the information carried by the case reference relationship to optimise the model. Extensive experiments have been conducted on two benchmark datasets, which demonstrate the state-of-the-art performance of CaseLink. The code has been released on https://github.com/yanran-tang/CaseLink.

Unicron: Economizing Self-Healing LLM Training at Scale

Training large-scale language models is increasingly critical in various domains, but it is hindered by frequent failures, leading to significant time and economic costs. Current failure recovery methods in cloud-based settings inadequately address the diverse and complex scenarios that arise, focusing narrowly on erasing downtime for individual tasks without considering the overall cost impact on a cluster. We introduce Unicron, a workload manager designed for efficient self-healing in large-scale language model training. Unicron optimizes the training process by minimizing failure-related costs across multiple concurrent tasks within a cluster. Its key features include in-band error detection for real-time error identification without extra overhead, a dynamic cost-aware plan generation mechanism for optimal reconfiguration, and an efficient transition strategy to reduce downtime during state changes. Deployed on a 128-GPU distributed cluster, Unicron demonstrates up to a 1.9x improvement in training efficiency over state-of-the-art methods, significantly reducing failure recovery costs and enhancing the reliability of large-scale language model training.