Xiaowen Jiang

Xiaowen Jiang, Sebastian U. Stich

The recently proposed stochastic Polyak stepsize (SPS) and stochastic line-search (SLS) for SGD have shown remarkable effectiveness when training over-parameterized models. However, in non-interpolation settings, both algorithms only guarantee convergence to a neighborhood of a solution which may result in a worse output than the initial guess. While artificially decreasing the adaptive stepsize has been proposed to address this issue (Orvieto et al. [2022]), this approach results in slower convergence rates for convex and over-parameterized models. In this work, we make two contributions: Firstly, we propose two new variants of SPS and SLS, called AdaSPS and AdaSLS, which guarantee convergence in non-interpolation settings and maintain sub-linear and linear convergence rates for convex and strongly convex functions when training over-parameterized models. AdaSLS requires no knowledge of problem-dependent parameters, and AdaSPS requires only a lower bound of the optimal function value as input. Secondly, we equip AdaSPS and AdaSLS with a novel variance reduction technique and obtain algorithms that require $\smash{\widetilde{\mathcal{O}}}(n+1/ε)$ gradient evaluations to achieve an $\mathcal{O}(ε)$-suboptimality for convex functions, which improves upon the slower $\mathcal{O}(1/ε^2)$ rates of AdaSPS and AdaSLS without variance reduction in the non-interpolation regimes. Moreover, our result matches the fast rates of AdaSVRG but removes the inner-outer-loop structure, which is easier to implement and analyze. Finally, numerical experiments on synthetic and real datasets validate our theory and demonstrate the effectiveness and robustness of our algorithms.

Xiaowen Jiang, Valerio Cambareri, Gianluca Agresti et al.

Sparse active illumination enables precise time-of-flight depth sensing as it maximizes signal-to-noise ratio for low power budgets. However, depth completion is required to produce dense depth maps for 3D perception. We address this task with realistic illumination and sensor resolution constraints by simulating ToF datasets for indoor 3D perception with challenging sparsity levels. We propose a quantized convolutional encoder-decoder network for this task. Our model achieves optimal depth map quality by means of input pre-processing and carefully tuned training with a geometry-preserving loss function. We also achieve low memory footprint for weights and activations by means of mixed precision quantization-at-training techniques. The resulting quantized models are comparable to the state of the art in terms of quality, but they require very low GPU times and achieve up to 14-fold memory size reduction for the weights w.r.t. their floating point counterpart with minimal impact on quality metrics.

Ali Zindari, Xiaowen Jiang, Rotem Mulayoff et al.

Low-rank adaptation (LoRA) is a widely used parameter-efficient fine-tuning method, yet its learned correction is static: the same low-rank update is applied to every input. This input-agnostic approach creates an inevitable compromise between adapting to the fine-tuning distribution and preserving pre-trained behavior on inputs outside that distribution, contributing to catastrophic forgetting. We introduce DISeL (Dynamic Input-Sensitive LoRA), which augments LoRA modules with lightweight input-dependent gates over individual rank-one components. The gating mechanism is designed to preserve the pre-trained model's behavior by default, while training learns to activate selected components that reduce the fine-tuning loss. DISeL adds only a small number of parameters and preserves the low-rank structure. Across RoBERTa on GLUE, and Llama and Mistral models fine-tuned for mathematical reasoning and code generation, DISeL reduces forgetting relative to LoRA and related variants while maintaining competitive fine-tuning accuracy. In addition, the learned gate activations provide an interpretable diagnostic view of which layers and rank components are most activated during fine-tuning, giving insight into where task-specific adaptation is concentrated. Code available at https://github.com/alizindari/DISeL .

Xiaowen Jiang, Anton Rodomanov, Sebastian U. Stich

In developing efficient optimization algorithms, it is crucial to account for communication constraints -- a significant challenge in modern Federated Learning. The best-known communication complexity among non-accelerated algorithms is achieved by DANE, a distributed proximal-point algorithm that solves local subproblems at each iteration and that can exploit second-order similarity among individual functions. However, to achieve such communication efficiency, the algorithm requires solving local subproblems sufficiently accurately resulting in slightly sub-optimal local complexity. Inspired by the hybrid-projection proximal-point method, in this work, we propose a novel distributed algorithm S-DANE. Compared to DANE, this method uses an auxiliary sequence of prox-centers while maintaining the same deterministic communication complexity. Moreover, the accuracy condition for solving the subproblem is milder, leading to enhanced local computation efficiency. Furthermore, S-DANE supports partial client participation and arbitrary stochastic local solvers, making it attractive in practice. We further accelerate S-DANE and show that the resulting algorithm achieves the best-known communication complexity among all existing methods for distributed convex optimization while still enjoying good local computation efficiency as S-DANE. Finally, we propose adaptive variants of both methods using line search, obtaining the first provably efficient adaptive algorithms that could exploit local second-order similarity without the prior knowledge of any parameters.

Haitian Jiang, Dongliang Xiong, Xiaowen Jiang et al.

Halftoning aims to reproduce a continuous-tone image with pixels whose intensities are constrained to two discrete levels. This technique has been deployed on every printer, and the majority of them adopt fast methods (e.g., ordered dithering, error diffusion) that fail to render structural details, which determine halftone's quality. Other prior methods of pursuing visual pleasure by searching for the optimal halftone solution, on the contrary, suffer from their high computational cost. In this paper, we propose a fast and structure-aware halftoning method via a data-driven approach. Specifically, we formulate halftoning as a reinforcement learning problem, in which each binary pixel's value is regarded as an action chosen by a virtual agent with a shared fully convolutional neural network (CNN) policy. In the offline phase, an effective gradient estimator is utilized to train the agents in producing high-quality halftones in one action step. Then, halftones can be generated online by one fast CNN inference. Besides, we propose a novel anisotropy suppressing loss function, which brings the desirable blue-noise property. Finally, we find that optimizing SSIM could result in holes in flat areas, which can be avoided by weighting the metric with the contone's contrast map. Experiments show that our framework can effectively train a light-weight CNN, which is 15x faster than previous structure-aware methods, to generate blue-noise halftones with satisfactory visual quality. We also present a prototype of deep multitoning to demonstrate the extensibility of our method.

Haitian Jiang, Dongliang Xiong, Xiaowen Jiang et al.

Deep neural networks have recently succeeded in digital halftoning using vanilla convolutional layers with high parallelism. However, existing deep methods fail to generate halftones with a satisfying blue-noise property and require complex training schemes. In this paper, we propose a halftoning method based on multi-agent deep reinforcement learning, called HALFTONERS, which learns a shared policy to generate high-quality halftone images. Specifically, we view the decision of each binary pixel value as an action of a virtual agent, whose policy is trained by a low-variance policy gradient. Moreover, the blue-noise property is achieved by a novel anisotropy suppressing loss function. Experiments show that our halftoning method produces high-quality halftones while staying relatively fast.

Radu-Alexandru Dragomir, Xiaowen Jiang, Bonan Sun et al.

Without randomization, escaping the saddle points of $f \colon \mathbb{R}^d \to \mathbb{R}$ requires at least $Ω(d)$ pieces of information about $f$ (values, gradients, Hessian-vector products). With randomization, this can be reduced to a polylogarithmic dependence in $d$. The prototypical algorithm to that effect is perturbed gradient descent (PGD): through sustained jitter, it reliably escapes strict saddle points. However, it also never settles: there is no convergence. What is more, PGD requires precise tuning based on Lipschitz constants and a preset target accuracy. To improve on this, we modify the time-tested trust-region method with truncated conjugate gradients (TR-tCG). Specifically, we randomize the initialization of tCG (the subproblem solver), and we prove that tCG automatically amplifies the randomization near saddles (to escape) and absorbs it near local minimizers (to converge). Saddle escape happens over several iterations. Accordingly, our analysis is multi-step, with several novelties. The proposed algorithm is practical: it essentially tracks the good behavior of TR-tCG, with three minute modifications and a single new hyperparameter (the noise scale $σ$). We provide an implementation and numerical experiments.

Xiaowen Jiang, Andrei Semenov, Sebastian U. Stich

While spectral-based optimizers like Muon operate directly on the spectrum of updates, standard adaptive methods such as AdamW do not account for the global spectral structure of weights and gradients, leaving them vulnerable to two empirical issues in large language model (LLM) training: (i) the optimizer updates can have large spectral norms, potentially destabilizing training and degrading generalization; (ii) stochastic gradient noise can exhibit sparse spectral spikes, with a few dominant singular values much larger than the rest. We propose SPECTRA, a general framework addressing these by (i) post-spectral clipping of updates to enforce spectral-norm constraints; (ii) optional pre-spectral clipping of gradients to suppress spectral noise spikes. We prove that post-clipping constitutes a Composite Frank-Wolfe method with spectral-norm constraints and weight regularization, recovering Frobenius and $\ell_{\infty}$-norm regularization with SGD-based and sign-based methods. We further analyze how pre-clipping mitigates sparse spectral spikes. We propose efficient soft spectral clipping via Newton-Schulz iterations, avoiding expensive SVD. Experiments on LLM pretraining show SPECTRA uniformly improves validation loss for various optimizers, including AdamW, Signum, and AdEMAMix, with the best-performing variants achieving state-of-the-art results. Models trained with SPECTRA exhibit smaller weight norms, confirming the link between spectral clipping and regularization.

Xiaowen Jiang, Anton Rodomanov, Sebastian U. Stich

Federated learning is a distributed optimization paradigm that allows training machine learning models across decentralized devices while keeping the data localized. The standard method, FedAvg, suffers from client drift which can hamper performance and increase communication costs over centralized methods. Previous works proposed various strategies to mitigate drift, yet none have shown uniformly improved communication-computation trade-offs over vanilla gradient descent. In this work, we revisit DANE, an established method in distributed optimization. We show that (i) DANE can achieve the desired communication reduction under Hessian similarity constraints. Furthermore, (ii) we present an extension, DANE+, which supports arbitrary inexact local solvers and has more freedom to choose how to aggregate the local updates. We propose (iii) a novel method, FedRed, which has improved local computational complexity and retains the same communication complexity compared to DANE/DANE+. This is achieved by using doubly regularized drift correction.

Xiaowen Jiang, Anton Rodomanov, Sebastian U. Stich

Different federated optimization algorithms typically employ distinct client-selection strategies: some methods communicate only with a randomly sampled subset of clients at each round, while others need to periodically communicate with all clients or use a hybrid scheme that combines both strategies. However, existing metrics for comparing optimization methods typically do not distinguish between these strategies, which often incur different communication costs in practice. To address this disparity, we introduce a simple and natural model of federated optimization that quantifies communication and local computation complexities. This new model allows for several commonly used client-selection strategies and explicitly associates each with a distinct cost. Within this setting, we propose a new algorithm that achieves the best-known communication and local complexities among existing federated optimization methods for non-convex optimization. This algorithm is based on the inexact composite gradient method with a carefully constructed gradient estimator and a special procedure for solving the auxiliary subproblem at each iteration. The gradient estimator is based on SAGA, a popular variance-reduced gradient estimator. We first derive a new variance bound for it, showing that SAGA can exploit functional similarity. We then introduce the Recursive-Gradient technique as a general way to potentially improve the error bound of a given conditionally unbiased gradient estimator, including both SAGA and SVRG. By applying this technique to SAGA, we obtain a new estimator, RG-SAGA, which has an improved error bound compared to the original one.

Yuki Takezawa, Anastasia Koloskova, Xiaowen Jiang et al.

Recently, a new optimization method based on the linear minimization oracle (LMO), called Muon, has been attracting increasing attention since it can train neural networks faster than existing adaptive optimization methods, such as Adam. In this paper, we study how Muon can be utilized in federated learning. We first show that straightforwardly using Muon as the local optimizer of FedAvg does not converge to the stationary point since the LMO is a biased operator. We then propose FedMuon which can mitigate this issue. We also analyze how solving the LMO approximately affects the convergence rate and find that, surprisingly, FedMuon can converge for any number of Newton-Schulz iterations, while it can converge faster as we solve the LMO more accurately. Through experiments, we demonstrated that FedMuon can outperform the state-of-the-art federated learning methods.