Rethinking Model Re-Basin and Linear Mode Connectivity

• URL: http://arxiv.org/abs/2402.05966v2
• Date: Tue, 9 Jul 2024 09:23:25 GMT
• Title: Rethinking Model Re-Basin and Linear Mode Connectivity
• Authors: Xingyu Qu, Samuel Horvath,
• Abstract summary: We decompose re-normalization into rescaling and reshift, uncovering that rescaling plays a crucial role in re-normalization. We identify that the merged model suffers from the issue of activation collapse and magnitude collapse. We propose a new perspective to unify the re-basin and pruning, under which a lightweight yet effective post-pruning technique is derived.
• Score: 1.1510009152620668
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent studies suggest that with sufficiently wide models, most SGD solutions can, up to permutation, converge into the same basin. This phenomenon, known as the model re-basin regime, has significant implications for model averaging by ensuring the linear mode connectivity. However, current re-basin strategies are ineffective in many scenarios due to a lack of comprehensive understanding of underlying mechanisms. Addressing this gap, this paper provides novel insights into understanding and improving the standard practice. Firstly, we decompose re-normalization into rescaling and reshift, uncovering that rescaling plays a crucial role in re-normalization while re-basin performance is sensitive to shifts in model activation. The finding calls for a more nuanced handling of the activation shift. Secondly, we identify that the merged model suffers from the issue of activation collapse and magnitude collapse. Varying the learning rate, weight decay, and initialization method can mitigate the issues and improve model performance. Lastly, we propose a new perspective to unify the re-basin and pruning, under which a lightweight yet effective post-pruning technique is derived, which can significantly improve the model performance after pruning. Our implementation is available at https://github.com/XingyuQu/rethink-re-basin.

Related papers

• ConsistentFeature: A Plug-and-Play Component for Neural Network Regularization [0.32885740436059047]
  Over- parameterized neural network models often lead to significant performance discrepancies between training and test sets. We introduce a simple perspective on overfitting: models learn different representations in different i.i.d. datasets. We propose an adaptive method, ConsistentFeature, that regularizes the model by constraining feature differences across random subsets of the same training set.
  arXiv  Detail & Related papers  (2024-12-02T13:21:31Z)
• SMILE: Zero-Shot Sparse Mixture of Low-Rank Experts Construction From Pre-Trained Foundation Models [85.67096251281191]
  We present an innovative approach to model fusion called zero-shot Sparse MIxture of Low-rank Experts (SMILE) construction. SMILE allows for the upscaling of source models into an MoE model without extra data or further training. We conduct extensive experiments across diverse scenarios, such as image classification and text generation tasks, using full fine-tuning and LoRA fine-tuning.
  arXiv  Detail & Related papers  (2024-08-19T17:32:15Z)
• Minusformer: Improving Time Series Forecasting by Progressively Learning Residuals [14.741951369068877]
  We find that ubiquitous time series (TS) forecasting models are prone to severe overfitting. We introduce a dual-stream and subtraction mechanism, which is a deep Boosting ensemble learning method. The proposed method outperform existing state-of-the-art methods, yielding an average performance improvement of 11.9% across various datasets.
  arXiv  Detail & Related papers  (2024-02-04T03:54:31Z)
• JoReS-Diff: Joint Retinex and Semantic Priors in Diffusion Model for Low-light Image Enhancement [69.6035373784027]
  Low-light image enhancement (LLIE) has achieved promising performance by employing conditional diffusion models. Previous methods may neglect the importance of a sufficient formulation of task-specific condition strategy. We propose JoReS-Diff, a novel approach that incorporates Retinex- and semantic-based priors as the additional pre-processing condition.
  arXiv  Detail & Related papers  (2023-12-20T08:05:57Z)
• AdaMerging: Adaptive Model Merging for Multi-Task Learning [68.75885518081357]
  This paper introduces an innovative technique called Adaptive Model Merging (AdaMerging) It aims to autonomously learn the coefficients for model merging, either in a task-wise or layer-wise manner, without relying on the original training data. Compared to the current state-of-the-art task arithmetic merging scheme, AdaMerging showcases a remarkable 11% improvement in performance.
  arXiv  Detail & Related papers  (2023-10-04T04:26:33Z)
• Sparse Model Soups: A Recipe for Improved Pruning via Model Averaging [24.64264715041198]
  We introduce Sparse Model Soups (SMS), a novel method for merging sparse models by initiating each prune-retrain cycle with the averaged model from the previous phase. SMS preserves sparsity, exploits sparse network benefits, is modular and fully parallelizable, and substantially improves IMP's performance.
  arXiv  Detail & Related papers  (2023-06-29T08:49:41Z)
• Exploiting Diffusion Prior for Real-World Image Super-Resolution [75.5898357277047]
  We present a novel approach to leverage prior knowledge encapsulated in pre-trained text-to-image diffusion models for blind super-resolution. By employing our time-aware encoder, we can achieve promising restoration results without altering the pre-trained synthesis model.
  arXiv  Detail & Related papers  (2023-05-11T17:55:25Z)
• Latent Autoregressive Source Separation [5.871054749661012]
  This paper introduces vector-quantized Latent Autoregressive Source Separation (i.e., de-mixing an input signal into its constituent sources) without requiring additional gradient-based optimization or modifications of existing models. Our separation method relies on the Bayesian formulation in which the autoregressive models are the priors, and a discrete (non-parametric) likelihood function is constructed by performing frequency counts over latent sums of addend tokens.
  arXiv  Detail & Related papers  (2023-01-09T17:32:00Z)
• ClusterQ: Semantic Feature Distribution Alignment for Data-Free Quantization [111.12063632743013]
  We propose a new and effective data-free quantization method termed ClusterQ. To obtain high inter-class separability of semantic features, we cluster and align the feature distribution statistics. We also incorporate the intra-class variance to solve class-wise mode collapse.
  arXiv  Detail & Related papers  (2022-04-30T06:58:56Z)
• Adversarial and Contrastive Variational Autoencoder for Sequential Recommendation [25.37244686572865]
  We propose a novel method called Adversarial and Contrastive Variational Autoencoder (ACVAE) for sequential recommendation. We first introduce the adversarial training for sequence generation under the Adversarial Variational Bayes framework, which enables our model to generate high-quality latent variables. Besides, when encoding the sequence, we apply a recurrent and convolutional structure to capture global and local relationships in the sequence.
  arXiv  Detail & Related papers  (2021-03-19T09:01:14Z)
• Bayesian Attention Modules [65.52970388117923]
  We propose a scalable version of attention that is easy to implement and optimize. Our experiments show the proposed method brings consistent improvements over the corresponding baselines.
  arXiv  Detail & Related papers  (2020-10-20T20:30:55Z)

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.