How to Fill the Optimum Set? Population Gradient Descent with Harmless Diversity

Authors: Chengyue Gong, Lemeng Wu, Qiang Liu

ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate that our method can efficiently generate diverse solutions on multiple applications, e.g. text-to-image generation, text-to-mesh generation, molecular conformation generation and ensemble neural network training.
Researcher Affiliation Academia 1Department of Computer Science, University of Texas at Austin.
Pseudocode Yes Algorithm 1 Diversity-aware Gradient Descent (Fsum)
Open Source Code No The paper links to 'https://github.com/NVlabs/stylegan2-ada-pytorch' in the appendix, which is a third-party tool used in their experiments, not the open-source code for their proposed methodology. There is no explicit statement or link indicating the release of their own source code.
Open Datasets Yes We use Big GAN for Image Net image generation and Style GAN-v2 for high-resolution image generation.
Dataset Splits No The paper mentions training and testing on datasets but does not explicitly provide details on how the datasets were split into training, validation, and test sets, nor does it mention a specific validation set split.
Hardware Specification Yes Hours is measured on a NVIDIA Ge Force RTX3090 GPU.
Software Dependencies No The paper mentions software components like 'Adam (Kingma & Ba, 2014) optimizer', 'Big GAN', 'Style GAN-v2', and 'Res Net-56 models', but does not provide specific version numbers for any programming languages, libraries, or frameworks used.
Experiment Setup Yes We adopt gradient descent with a constant learning rate 5 10 4 and 1,000 iterations.