Realistic Evaluation of Model Merging for Compositional Generalization

• URL: http://arxiv.org/abs/2409.18314v1
• Date: Thu, 26 Sep 2024 21:44:20 GMT
• Title: Realistic Evaluation of Model Merging for Compositional Generalization
• Authors: Derek Tam, Yash Kant, Brian Lester, Igor Gilitschenski, Colin Raffel,
• Abstract summary: We evaluate the relative merits of different merging methods by evaluating them in a shared experimental setting. Specifically, our setting focuses on using merging for compositional generalization of capabilities in image classification, image generation, and natural language processing. We measure the computational costs of different merging methods as well as how they perform when scaling the number of models being merged.
• Score: 45.68423348264746
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Merging has become a widespread way to cheaply combine individual models into a single model that inherits their capabilities and attains better performance. This popularity has spurred rapid development of many new merging methods, which are typically validated in disparate experimental settings and frequently differ in the assumptions made about model architecture, data availability, and computational budget. In this work, we characterize the relative merits of different merging methods by evaluating them in a shared experimental setting and precisely identifying the practical requirements of each method. Specifically, our setting focuses on using merging for compositional generalization of capabilities in image classification, image generation, and natural language processing. Additionally, we measure the computational costs of different merging methods as well as how they perform when scaling the number of models being merged. Taken together, our results clarify the state of the field of model merging and provide a comprehensive and rigorous experimental setup to test new methods.

Related papers

• Collective Model Intelligence Requires Compatible Specialization [29.590052023903457]
  We show that as models specialize, the similarity in their feature space structure diminishes, hindering their capacity for collective use. We propose a new direction for achieving collective model intelligence through what we call compatible specialization.
  arXiv  Detail & Related papers  (2024-11-04T15:59:16Z)
• A Simple Approach to Unifying Diffusion-based Conditional Generation [63.389616350290595]
  We introduce a simple, unified framework to handle diverse conditional generation tasks. Our approach enables versatile capabilities via different inference-time sampling schemes. Our model supports additional capabilities like non-spatially aligned and coarse conditioning.
  arXiv  Detail & Related papers  (2024-10-15T09:41:43Z)
• Parameter Competition Balancing for Model Merging [13.66727853299506]
  PCB-Merging is a training-free technique that adjusts the coefficients of each parameter for effective model merging. PCB-Merging achieves substantial performance enhancements across multiple modalities, domains, model sizes, number of tasks, fine-tuning forms, and large language models.
  arXiv  Detail & Related papers  (2024-10-03T11:17:58Z)
• Localize-and-Stitch: Efficient Model Merging via Sparse Task Arithmetic [22.73746175315071]
  We introduce Localize-and-Stitch, a novel approach that merges models in a localized way. We demonstrate that our approach effectively locates sparse regions responsible for finetuned performance. Our algorithm also facilitates model compression and preserves pretrained knowledge.
  arXiv  Detail & Related papers  (2024-08-24T19:14:02Z)
• FusionBench: A Comprehensive Benchmark of Deep Model Fusion [78.80920533793595]
  Deep model fusion is a technique that unifies the predictions or parameters of several deep neural networks into a single model. FusionBench is the first comprehensive benchmark dedicated to deep model fusion.
  arXiv  Detail & Related papers  (2024-06-05T13:54:28Z)
• Model ensemble instead of prompt fusion: a sample-specific knowledge transfer method for few-shot prompt tuning [85.55727213502402]
  We focus on improving the few-shot performance of prompt tuning by transferring knowledge from soft prompts of source tasks. We propose Sample-specific Ensemble of Source Models (SESoM) SESoM learns to adjust the contribution of each source model for each target sample separately when ensembling source model outputs.
  arXiv  Detail & Related papers  (2022-10-23T01:33:16Z)
• HyperImpute: Generalized Iterative Imputation with Automatic Model Selection [77.86861638371926]
  We propose a generalized iterative imputation framework for adaptively and automatically configuring column-wise models. We provide a concrete implementation with out-of-the-box learners, simulators, and interfaces.
  arXiv  Detail & Related papers  (2022-06-15T19:10:35Z)
• Federated Learning Aggregation: New Robust Algorithms with Guarantees [63.96013144017572]
  Federated learning has been recently proposed for distributed model training at the edge. This paper presents a complete general mathematical convergence analysis to evaluate aggregation strategies in a federated learning framework. We derive novel aggregation algorithms which are able to modify their model architecture by differentiating client contributions according to the value of their losses.
  arXiv  Detail & Related papers  (2022-05-22T16:37:53Z)
• Split Modeling for High-Dimensional Logistic Regression [0.2676349883103404]
  A novel method is proposed to an ensemble logistic classification model briefly compiled. Our method learns how to exploit the bias-off resulting in excellent prediction accuracy. An open-source software library implementing the proposed method is discussed.
  arXiv  Detail & Related papers  (2021-02-17T05:57:26Z)
• Amortized Bayesian model comparison with evidential deep learning [0.12314765641075436]
  We propose a novel method for performing Bayesian model comparison using specialized deep learning architectures. Our method is purely simulation-based and circumvents the step of explicitly fitting all alternative models under consideration to each observed dataset. We show that our method achieves excellent results in terms of accuracy, calibration, and efficiency across the examples considered in this work.
  arXiv  Detail & Related papers  (2020-04-22T15:15:46Z)

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.