Gaussian Stochastic Weight Averaging for Bayesian Low-Rank Adaptation of Large Language Models

• URL: http://arxiv.org/abs/2405.03425v2
• Date: Sat, 20 Jul 2024 04:36:27 GMT
• Title: Gaussian Stochastic Weight Averaging for Bayesian Low-Rank Adaptation of Large Language Models
• Authors: Emre Onal, Klemens Flöge, Emma Caldwell, Arsen Sheverdin, Vincent Fortuin,
• Abstract summary: Fine-tuned Large Language Models (LLMs) often suffer from overconfidence and poor calibration. We propose a simple combination of Low-Rank Adaptation (LoRA) with Gaussian Weight Averaging (SWAG) We show that our method exhibits greater robustness against distribution shift, as reflected in its improved performance on out-of-distribution tasks.
• Score: 5.352221132808875
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Fine-tuned Large Language Models (LLMs) often suffer from overconfidence and poor calibration, particularly when fine-tuned on small datasets. To address these challenges, we propose a simple combination of Low-Rank Adaptation (LoRA) with Gaussian Stochastic Weight Averaging (SWAG), facilitating approximate Bayesian inference in LLMs. Through extensive testing across several Natural Language Processing (NLP) benchmarks, we demonstrate that our straightforward and computationally efficient approach improves model generalization and calibration competitively with comparable, more sophisticated methods for Bayesian inference in LLMs. We further show that our method exhibits greater robustness against distribution shift, as reflected in its improved performance on out-of-distribution tasks.

Related papers

• GIFT-SW: Gaussian noise Injected Fine-Tuning of Salient Weights for LLMs [51.02233412547456]
  We introduce a novel PEFT method, called Gaussian noise Injected Fine Tuning of Salient Weights (GIFT-SW) Our method updates only salient columns, while injecting Gaussian noise into non-salient ones. Experiments with LLaMA models demonstrate that GIFT-SW outperforms full fine-tuning and modern PEFT methods under the same computational budget.
  arXiv  Detail & Related papers  (2024-08-27T14:41:14Z)
• Regression-aware Inference with LLMs [52.764328080398805]
  We show that an inference strategy can be sub-optimal for common regression and scoring evaluation metrics. We propose alternate inference strategies that estimate the Bayes-optimal solution for regression and scoring metrics in closed-form from sampled responses.
  arXiv  Detail & Related papers  (2024-03-07T03:24:34Z)
• Calibrating Large Language Models with Sample Consistency [76.23956851098598]
  We explore the potential of deriving confidence from the distribution of multiple randomly sampled model generations, via three measures of consistency. Results show that consistency-based calibration methods outperform existing post-hoc approaches. We offer practical guidance on choosing suitable consistency metrics for calibration, tailored to the characteristics of various LMs.
  arXiv  Detail & Related papers  (2024-02-21T16:15:20Z)
• CATfOOD: Counterfactual Augmented Training for Improving Out-of-Domain Performance and Calibration [59.48235003469116]
  We show that data augmentation consistently enhances OOD performance. We also show that CF augmented models which are easier to calibrate also exhibit much lower entropy when assigning importance.
  arXiv  Detail & Related papers  (2023-09-14T16:16:40Z)
• Optimization of Annealed Importance Sampling Hyperparameters [77.34726150561087]
  Annealed Importance Sampling (AIS) is a popular algorithm used to estimates the intractable marginal likelihood of deep generative models. We present a parameteric AIS process with flexible intermediary distributions and optimize the bridging distributions to use fewer number of steps for sampling. We assess the performance of our optimized AIS for marginal likelihood estimation of deep generative models and compare it to other estimators.
  arXiv  Detail & Related papers  (2022-09-27T07:58:25Z)
• Generalised Gaussian Process Latent Variable Models (GPLVM) with Stochastic Variational Inference [9.468270453795409]
  We study the doubly formulation of the BayesianVM model amenable with minibatch training. We show how this framework is compatible with different latent variable formulations and perform experiments to compare a suite of models. We demonstrate how we can train in the presence of massively missing data and obtain high-fidelity reconstructions.
  arXiv  Detail & Related papers  (2022-02-25T21:21:51Z)
• Scalable Cross Validation Losses for Gaussian Process Models [22.204619587725208]
  We use Polya-Gamma auxiliary variables and variational inference to accommodate binary and multi-class classification. We find that our method offers fast training and excellent predictive performance.
  arXiv  Detail & Related papers  (2021-05-24T21:01:47Z)
• Scalable Control Variates for Monte Carlo Methods via Stochastic Optimization [62.47170258504037]
  This paper presents a framework that encompasses and generalizes existing approaches that use controls, kernels and neural networks. Novel theoretical results are presented to provide insight into the variance reduction that can be achieved, and an empirical assessment, including applications to Bayesian inference, is provided in support.
  arXiv  Detail & Related papers  (2020-06-12T22:03:25Z)
• Fitting Laplacian Regularized Stratified Gaussian Models [0.0]
  We consider the problem of jointly estimating multiple related zero-mean Gaussian distributions from data. We propose a distributed method that scales to large problems, and illustrate the efficacy of the method with examples in finance, radar signal processing, and weather forecasting.
  arXiv  Detail & Related papers  (2020-05-04T18:00:59Z)
• Sparse Gaussian Processes Revisited: Bayesian Approaches to Inducing-Variable Approximations [27.43948386608]
  Variational inference techniques based on inducing variables provide an elegant framework for scalable estimation in Gaussian process (GP) models. In this work we challenge the common wisdom that optimizing the inducing inputs in variational framework yields optimal performance.
  arXiv  Detail & Related papers  (2020-03-06T08:53:18Z)
• Bayesian Neural Networks With Maximum Mean Discrepancy Regularization [13.97417198693205]
  We show that our BNNs achieve higher accuracy on multiple benchmarks, including several image classification tasks. We also provide a new formulation for estimating the uncertainty on a given prediction, showing it performs in a more robust fashion against adversarial attacks.
  arXiv  Detail & Related papers  (2020-03-02T14:54:48Z)

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.