Spectrum: Targeted Training on Signal to Noise Ratio

• URL: http://arxiv.org/abs/2406.06623v1
• Date: Fri, 7 Jun 2024 21:20:57 GMT
• Title: Spectrum: Targeted Training on Signal to Noise Ratio
• Abstract summary: We present Spectrum, a method that accelerates LLM training by selectively targeting layer modules based on their signal-to-noise ratio (SNR) Our approach, which utilizes an algorithm to compute module SNRs prior to training, has shown to effectively match the performance of full fine-tuning while reducing GPU memory usage.
• Score: 39.74810517186403
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Efficiently post-training large language models remains a challenging task due to the vast computational resources required. We present Spectrum, a method that accelerates LLM training by selectively targeting layer modules based on their signal-to-noise ratio (SNR), and freezing the remaining modules. Our approach, which utilizes an algorithm to compute module SNRs prior to training, has shown to effectively match the performance of full fine-tuning while reducing GPU memory usage. Experiments comparing Spectrum to existing methods such as QLoRA demonstrate its effectiveness in terms of model quality and VRAM efficiency in distributed environments.

Related papers

• Revealing Modular Gradient Noise Imbalance in LLMs: Calibrating Adam via Signal-to-Noise Ratio [45.81412107070137]
  We introduce textbfModule-wise Learning Rate Scaling via SNR (MoLS).<n>MoLS estimates module-level SNRs to scale Adam updates, allowing automated module-wise learning rate allocation without manual tuning.<n> Empirical results through multiple LLM training benchmarks demonstrate that MoLS improves convergence speed and generalization, achieving performance comparable to carefully tuned module-specific learning rates.
  arXiv  Detail & Related papers  (2026-05-07T07:32:27Z)
• Timestep-Aware Block Masking for Efficient Diffusion Model Inference [48.764748095994996]
  Diffusion Probabilistic Models (DPMs) have achieved great success in image generation but suffer from high inference latency.<n>We propose a novel framework to optimize the computational graph of pre-trained DPMs on a per-timestep basis.
  arXiv  Detail & Related papers  (2026-03-20T13:35:03Z)
• Language Ranker: A Lightweight Ranking framework for LLM Decoding [70.01564145836129]
  This paper conceptualizes the decoding process as analogous to the ranking stage in recommendation pipelines.<n>Motivated by this insight, we propose Language Ranker, a novel framework that introduces a lightweight module to rerank candidate responses.<n> Experiments show that Language Ranker achieves performance comparable to large-scale reward models, while requiring only 0.5M additional parameters.
  arXiv  Detail & Related papers  (2025-10-23T17:56:46Z)
• A Measurement Report Data-Driven Framework for Localized Statistical Channel Modeling [25.91473621688449]
  Localized statistical channel modeling (LSCM) is crucial for effective performance evaluation in digital twin-assisted network optimization.<n>Existing methods rely heavily on drive test data with high collection costs and limited spatial coverage.<n>We propose a measurement report (MR) data-driven framework for LSCM, exploiting the low-cost and extensive collection of MR data.
  arXiv  Detail & Related papers  (2025-09-16T10:59:22Z)
• When marine radar target detection meets pretrained large language models [19.91452033424555]
  We propose a framework that integrates feature preprocessing with large language models (LLMs)<n>Our preprocessing module tokenizes radar sequence features, applies a patch selection algorithm to filter out uninformative segments, and projects the selected patches into embeddings compatible with the feature space of pre-trained LLMs.<n> Experiments on measured datasets demonstrate that the proposed method significantly outperforms the state-of-the-art baselines on supervised learning tests.
  arXiv  Detail & Related papers  (2025-09-15T16:38:13Z)
• Small-Scale-Fading-Aware Resource Allocation in Wireless Federated Learning [27.931985523249352]
  This paper proposes a small-scale-fading-aware resource allocation strategy using a multi-agent reinforcement learning (MARL) framework.<n>We establish a one-step convergence bound of the FL algorithm and formulate the resource allocation problem as a decentralized partially observable Markov decision process.<n>In our framework, each client serves as an agent that dynamically determines spectrum and power allocations within each coherence time slot.
  arXiv  Detail & Related papers  (2025-05-06T13:41:59Z)
• Spatial-Spectral Diffusion Contrastive Representation Network for Hyperspectral Image Classification [8.600534616819333]
  This paper presents a Spatial-Spectral Diffusion Contrastive Representation Network (DiffCRN) DiffCRN is based on denoising diffusion probabilistic model (DDPM) combined with contrastive learning (CL) for hyperspectral images classification. Experiments conducted on widely used four HSI datasets demonstrate the improved performance of the proposed DiffCRN.
  arXiv  Detail & Related papers  (2025-02-27T02:34:23Z)
• Less is More: Extreme Gradient Boost Rank-1 Adaption for Efficient Finetuning of LLMs [75.11449420928139]
  Fine-tuning Large Language Models (LLMs) has become a crucial technique for adapting pre-trained models to downstream tasks. Low-Rank Adaptation (LoRA) has emerged as a promising solution, but there exists a gap between the practical performance of low-rank adaptations and its theoretical optimum. We propose eXtreme Gradient Boosting LoRA, a novel framework that bridges this gap by leveraging the power of ensemble learning.
  arXiv  Detail & Related papers  (2024-10-25T17:07:13Z)
• R-SFLLM: Jamming Resilient Framework for Split Federated Learning with Large Language Models [83.77114091471822]
  Split federated learning (SFL) is a compute-efficient paradigm in distributed machine learning (ML) A challenge in SFL, particularly when deployed over wireless channels, is the susceptibility of transmitted model parameters to adversarial jamming. This is particularly pronounced for word embedding parameters in large language models (LLMs), which are crucial for language understanding. A physical layer framework is developed for resilient SFL with LLMs (R-SFLLM) over wireless networks.
  arXiv  Detail & Related papers  (2024-07-16T12:21:29Z)
• Tender: Accelerating Large Language Models via Tensor Decomposition and Runtime Requantization [0.6445087473595953]
  Large language models (LLMs) demonstrate outstanding performance in various tasks in machine learning. deploying LLM inference poses challenges due to the high compute and memory requirements. We present Tender, an algorithm-hardware co-design solution that enables efficient deployment of LLM inference at low precision.
  arXiv  Detail & Related papers  (2024-06-16T09:51:55Z)
• Adaptive Layer Splitting for Wireless LLM Inference in Edge Computing: A Model-Based Reinforcement Learning Approach [18.153641696306707]
  This study introduces a framework taking inspiration from model-based reinforcement learning (MBRL) to determine the optimal splitting point across the edge and user equipment (UE) By incorporating a reward surrogate model, our approach significantly reduces the computational cost of frequent performance evaluations.
  arXiv  Detail & Related papers  (2024-06-03T09:41:42Z)
• Convergence Rate Maximization for Split Learning-based Control of EMG Prosthetic Devices [2.432653781859026]
  Split Learning (SL) is a promising Distributed Learning approach in electromyography (EMG) based prosthetic control. This paper presents an algorithm for optimal cut layer selection in terms of maximizing the convergence rate of the model.
  arXiv  Detail & Related papers  (2024-01-06T15:05:49Z)
• Federated Conditional Stochastic Optimization [110.513884892319]
  Conditional optimization has found in a wide range of machine learning tasks, such as in-variant learning tasks, AUPRC, andAML. This paper proposes algorithms for distributed federated learning.
  arXiv  Detail & Related papers  (2023-10-04T01:47:37Z)
• Online Network Source Optimization with Graph-Kernel MAB [62.6067511147939]
  We propose Grab-UCB, a graph- kernel multi-arms bandit algorithm to learn online the optimal source placement in large scale networks. We describe the network processes with an adaptive graph dictionary model, which typically leads to sparse spectral representations. We derive the performance guarantees that depend on network parameters, which further influence the learning curve of the sequential decision strategy.
  arXiv  Detail & Related papers  (2023-07-07T15:03:42Z)
• Loop Unrolled Shallow Equilibrium Regularizer (LUSER) -- A Memory-Efficient Inverse Problem Solver [26.87738024952936]
  In inverse problems we aim to reconstruct some underlying signal of interest from potentially corrupted and often ill-posed measurements. We propose an LU algorithm with shallow equilibrium regularizers (L) These implicit models are as expressive as deeper convolutional networks, but far more memory efficient during training.
  arXiv  Detail & Related papers  (2022-10-10T19:50:37Z)
• Approximating the Ideal Observer for joint signal detection and localization tasks by use of supervised learning methods [15.226790614827193]
  Medical imaging systems are commonly assessed and optimized by use of objective measures of image quality (IQ) The Ideal Observer (IO) performance has been advocated to provide a figure-of-merit for use in assessing and optimizing imaging systems. In this paper, the ability of supervised learning-based methods to approximate the IO for joint signal detection and localization tasks is explored.
  arXiv  Detail & Related papers  (2020-05-29T22:53:45Z)
• Optimization-driven Deep Reinforcement Learning for Robust Beamforming in IRS-assisted Wireless Communications [54.610318402371185]
  Intelligent reflecting surface (IRS) is a promising technology to assist downlink information transmissions from a multi-antenna access point (AP) to a receiver. We minimize the AP's transmit power by a joint optimization of the AP's active beamforming and the IRS's passive beamforming. We propose a deep reinforcement learning (DRL) approach that can adapt the beamforming strategies from past experiences.
  arXiv  Detail & Related papers  (2020-05-25T01:42:55Z)

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.