Efficient Hyperparameter Search for Non-Stationary Model Training

• URL: http://arxiv.org/abs/2512.01258v1
• Date: Mon, 01 Dec 2025 04:06:24 GMT
• Title: Efficient Hyperparameter Search for Non-Stationary Model Training
• Abstract summary: We introduce a two-stage paradigm to reduce the cost of model training for online learning systems.<n>Our core insight is that focusing on accurate identification in the first stage, rather than achieving peak performance, allows for aggressive cost-saving measures.<n>We validate our framework's effectiveness through a dual evaluation: first on the Criteo 1TB dataset, the largest suitable public benchmark, and second on an industrial advertising system operating at a scale two orders of magnitude larger.
• Score: 11.55466013293614
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Online learning is the cornerstone of applications like recommendation and advertising systems, where models continuously adapt to shifting data distributions. Model training for such systems is remarkably expensive, a cost that multiplies during hyperparameter search. We introduce a two-stage paradigm to reduce this cost: (1) efficiently identifying the most promising configurations, and then (2) training only these selected candidates to their full potential. Our core insight is that focusing on accurate identification in the first stage, rather than achieving peak performance, allows for aggressive cost-saving measures. We develop novel data reduction and prediction strategies that specifically overcome the challenges of sequential, non-stationary data not addressed by conventional hyperparameter optimization. We validate our framework's effectiveness through a dual evaluation: first on the Criteo 1TB dataset, the largest suitable public benchmark, and second on an industrial advertising system operating at a scale two orders of magnitude larger. Our methods reduce the total hyperparameter search cost by up to 10$\times$ on the public benchmark and deliver significant, validated efficiency gains in the industrial setting.

Related papers

• An `Inverse' Experimental Framework to Estimate Market Efficiency [0.8594140167290097]
  Digital marketplaces processing billions of dollars annually represent critical infrastructure in sociotechnical ecosystems.<n>We develop and train several models, including linear regressions and gradient boosting trees, leveraging quantile-based input from the underlying supply-demand model.
  arXiv  Detail & Related papers  (2026-04-20T11:52:54Z)
• Cheap Thrills: Effective Amortized Optimization Using Inexpensive Labels [20.00525916892172]
  We propose "cheap imperfect labels," then perform pretraining, and refine the model through self-supervised learning to improve overall performance.<n>Our theoretical analysis and empirically-based criterion show that labeled data only need place the model within a basin of attraction.<n>We show it yields faster convergence; improved accuracy; high-quality optimality; and up to 59x reductions in total offline cost.
  arXiv  Detail & Related papers  (2026-03-05T18:58:39Z)
• Cost-Aware Retrieval-Augmentation Reasoning Models with Adaptive Retrieval Depth [18.05278637533445]
  We propose a retrieval-augmented reasoning model that dynamically adjusts the length of the retrieved document list.<n>We develop a cost-aware advantage function for training of efficient retrieval-augmented reasoning models through reinforcement learning.<n>We evaluate our approach on seven public question answering datasets and demonstrate significant efficiency gains, without compromising effectiveness.
  arXiv  Detail & Related papers  (2025-10-17T15:04:03Z)
• Thinking Augmented Pre-training [88.04395622064708]
  Thinking augmented Pre-Training is a universal methodology that augments text with automatically generated thinking trajectories.<n>This paper introduces a simple and scalable approach to improve the data efficiency of large language model (LLM) training by augmenting existing text data with thinking trajectories.
  arXiv  Detail & Related papers  (2025-09-24T14:45:13Z)
• Towards High Data Efficiency in Reinforcement Learning with Verifiable Reward [54.708851958671794]
  We propose a Data-Efficient Policy Optimization pipeline that combines optimized strategies for both offline and online data selection.<n>In offline phase, we curate a high-quality subset of training samples based on diversity, influence, and appropriate difficulty.<n>During online RLVR training, we introduce a sample-level explorability metric to dynamically filter samples with low exploration potential.
  arXiv  Detail & Related papers  (2025-09-01T10:04:20Z)
• Scalable Parameter and Memory Efficient Pretraining for LLM: Recent Algorithmic Advances and Benchmarking [26.80161478380058]
  Large language models (LLMs) have grown at an unprecedented rate, with some recent models containing trillions of parameters.<n>This growth is accompanied by substantial computational challenges, particularly regarding the memory and compute resources required for training and fine-tuning.<n>Motivated by this issue, we aim to address the following questions: Can parameter- or memory-efficient methods enhance pre-training efficiency while achieving performance comparable to full-model training?
  arXiv  Detail & Related papers  (2025-05-28T22:51:43Z)
• Self-Steering Optimization: Autonomous Preference Optimization for Large Language Models [79.84205827056907]
  We present Self-Steering Optimization ($SSO$), an algorithm that autonomously generates high-quality preference data.<n>$SSO$ employs a specialized optimization objective to build a data generator from the policy model itself, which is used to produce accurate and on-policy data.<n>Our evaluation shows that $SSO$ consistently outperforms baselines in human preference alignment and reward optimization.
  arXiv  Detail & Related papers  (2024-10-22T16:04:03Z)
• Cost-Effective Proxy Reward Model Construction with On-Policy and Active Learning [70.22819290458581]
  Reinforcement learning with human feedback (RLHF) is a widely adopted approach in current large language model pipelines. Our approach introduces two key innovations: (1) on-policy query to avoid OOD and imbalance issues in seed data, and (2) active learning to select the most informative data for preference queries.
  arXiv  Detail & Related papers  (2024-07-02T10:09:19Z)
• Out of Thin Air: Exploring Data-Free Adversarial Robustness Distillation [26.744403789694758]
  We propose Data-Free Adversarial Robustness Distillation (DFARD) to train small, easily deployable, robust models without relying on data. Inspired by human education, we design a plug-and-play Interactive Temperature Adjustment (ITA) strategy to improve the efficiency of knowledge transfer.
  arXiv  Detail & Related papers  (2023-03-21T06:10:47Z)
• VFed-SSD: Towards Practical Vertical Federated Advertising [53.08038962443853]
  We propose a semi-supervised split distillation framework VFed-SSD to alleviate the two limitations. Specifically, we develop a self-supervised task MatchedPair Detection (MPD) to exploit the vertically partitioned unlabeled data. Our framework provides an efficient federation-enhanced solution for real-time display advertising with minimal deploying cost and significant performance lift.
  arXiv  Detail & Related papers  (2022-05-31T17:45:30Z)
• Amortized Auto-Tuning: Cost-Efficient Transfer Optimization for Hyperparameter Recommendation [83.85021205445662]
  We propose an instantiation--amortized auto-tuning (AT2) to speed up tuning of machine learning models. We conduct a thorough analysis of the multi-task multi-fidelity Bayesian optimization framework, which leads to the best instantiation--amortized auto-tuning (AT2)
  arXiv  Detail & Related papers  (2021-06-17T00:01:18Z)
• Toward Optimal Probabilistic Active Learning Using a Bayesian Approach [4.380488084997317]
  Active learning aims at reducing the labeling costs by an efficient and effective allocation of costly labeling resources. By reformulating existing selection strategies within our proposed model, we can explain which aspects are not covered in current state-of-the-art.
  arXiv  Detail & Related papers  (2020-06-02T15:59:42Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.