Analysis of Quantized Models

Authors: Lu Hou, Ruiliang Zhang, James T. Kwok

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

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical experiments confirm the theoretical convergence results, and demonstrate that quantized networks can speed up training and have comparable performance as full-precision networks.
Researcher Affiliation Collaboration Lu Hou1, Ruiliang Zhang1,2, James T. Kwok1 1Department of Computer Science and Engineering Hong Kong University of Science and Technology Hong Kong {lhouab,jamesk}@cse.ust.hk 2Tu Simple ruiliang.zhang@tusimple.ai
Pseudocode No The paper contains mathematical derivations and descriptions of methods, but no formally labeled pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statement or link indicating that source code is open or publicly available for the described methodology.
Open Datasets Yes In this experiment, we follow (Wen et al., 2017) and use the same train/test split, data preprocessing, augmentation and distributed Tensorflow setup. ... We train the Alex Net on Image Net.
Dataset Splits Yes In this experiment, we follow (Wen et al., 2017) and use the same train/test split, data preprocessing, augmentation and distributed Tensorflow setup.
Hardware Specification Yes Speedup of Image Net training on a 16-node GPU cluster. Each node has 4 1080ti GPUs with one PCI switch.
Software Dependencies No The paper mentions 'distributed Tensorflow setup' and 'Adam is used as the optimizer' but does not specify version numbers for these software components.
Experiment Setup Yes The optimizer is RMSProp, and the learning rate is ηt = η/t, where η = 0.03. Training is terminated when the average training loss does not decrease for 5000 iterations. ... The learning rate is decayed from 0.0002 by a factor of 0.1 every 200 epochs.