Optimistic mirror descent in saddle-point problems: Going the extra (gradient) mile

Authors: Panayotis Mertikopoulos, Bruno Lecouat, Houssam Zenati, Chuan-Sheng Foo, Vijay Chandrasekhar, Georgios Piliouras

ICLR 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We also validate our analysis by numerical experiments in a wide array of GAN models (including Gaussian mixture models, and the Celeb A and CIFAR-10 datasets).
Researcher Affiliation Academia Panayotis Mertikopoulos Univ. Grenoble Alpes, CNRS, Inria, Grenoble INP, LIG 38000 Grenoble, France panayotis.mertikopoulos@imag.fr Bruno Lecouat, Houssam Zenati, Chuan-Sheng Foo, Vijay Chandrasekhar Institute for Infocomm Research, A*STAR 1 Fusionopolis Way, #21-01 Connexis (South Tower), Singapore {bruno_lecouat,houssam_zenati,foocs,vijay}@i2r.a-star.edu.sg Georgios Piliouras Singapore University of Technology and Design 8 Somapah Road, Singapore georgios@sutd.edu.sg
Pseudocode Yes Algorithm 1: mirror descent (MD) for saddle-point problems; Algorithm 2: optimistic mirror descent (OMD) for saddle-point problems; Algorithm 3: Adam with extra-gradient add-on (optimistic Adam)
Open Source Code No The paper does not provide a concrete access to source code (specific repository link, explicit code release statement, or code in supplementary materials) for the methodology described.
Open Datasets Yes We also validate our analysis by numerical experiments in a wide array of GAN models (including Gaussian mixture models, and the Celeb A and CIFAR-10 datasets).
Dataset Splits No The paper does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning into train/validation/test sets.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions training with RMSprop and Adam, but it does not provide specific ancillary software details, such as library names with version numbers, needed to replicate the experiment.
Experiment Setup Yes Table 2: Image experiments settings batch size = 64 Adam learning rate = 0.0001 Adam β1 = 0.0 Adam β2 = 0.9 max iterations = 200000 WGAN-GP λ = 1.0 WGAN-GP ndis = 1 GAN objective = WGAN-GP Optimizer = extra-Adam or Adam