paper-with-me

Papers

How much progress have we made in neural network training? A New Evaluation Protocol for Benchmarking Optimizers

2020-10-19 · Yuanhao Xiong, Xuanqing Liu, Li-Cheng Lan, Yang You, Si Si, Cho-Jui Hsieh

Many optimizers have been proposed for training deep neural networks, and they often have multiple hyperparameters, which make it tricky to benchmark their performance. In this work, we propose a new benchmarking protocol to evaluate both end-to-end efficiency (training a model from scratch without knowing the best hyperparameter) and data-addition training efficiency (the previously selected hyperparameters are used for periodically re-training the model with newly collected data). For end-to-end efficiency, unlike previous work that assumes random hyperparameter tuning, which over-emphasizes the tuning time, we propose to evaluate with a bandit hyperparameter tuning strategy. A human study is conducted to show that our evaluation protocol matches human tuning behavior better than the random search. For data-addition training, we propose a new protocol for assessing the hyperparameter sensitivity to data shift. We then apply the proposed benchmarking framework to 7 optimizers and various tasks, including computer vision, natural language processing, reinforcement learning, and graph mining. Our results show that there is no clear winner across all the tasks.

📄 PDF Abstract BibTeX arXiv:2010.09889

Code (0)

등록된 구현이 없습니다.

Tasks

BenchmarkingGraph Mining

Similar Papers 제목 키워드 기반

Measuring Fairness in Generative Models

2021-07-16 · Christopher T. H Teo, Ngai-Man Cheung

Deep generative models have made much progress in improving training stability and quality of generated data. Recently there has been increased interest in the fairness of deep-generated data. Fairness is important in ma…

Fairness

Towards energy-efficient Deep Learning: An overview of energy-efficient approaches along the Deep Learning Lifecycle

2023-02-05 · Vanessa Mehlin, Sigurd Schacht, Carsten Lanquillon

Deep Learning has enabled many advances in machine learning applications in the last few years. However, since current Deep Learning algorithms require much energy for computations, there are growing concerns about the a…

Deep Learning

Scaling Web Agent Training through Automatic Data Generation and Fine-grained Evaluation

2026-02-13 · Lajanugen Logeswaran, Jaekyeom Kim, Sungryull Sohn, Creighton Glasscock 외 arxiv

We present a scalable pipeline for automatically generating high-quality training data for web agents. In particular, a major challenge in identifying high-quality training instances is trajectory evaluation - quantifyin…

UniBench: Visual Reasoning Requires Rethinking Vision-Language Beyond Scaling

2024-08-09 · Haider Al-Tahan, Quentin Garrido, Randall Balestriero, Diane Bouchacourt 외

Significant research efforts have been made to scale and improve vision-language model (VLM) training approaches. Yet, with an ever-growing number of benchmarks, researchers are tasked with the heavy burden of implementi…

GPULanguage ModelingLanguage ModellingObject Recognition+1

How much progress have we made on RST discourse parsing? A replication study of recent results on the RST-DT

2017-09-01 · EMNLP 2017 9 · Mathieu Morey, Philippe Muller, Nicholas Asher

This article evaluates purported progress over the past years in RST discourse parsing. Several studies report a relative error reduction of 24 to 51{\%} on all metrics that authors attribute to the introduction of distr…

AttributeDiscourse Parsing