Sanity-Checking Pruning Methods: Random Tickets can Win the Jackpot
Authors: Jingtong Su, Yihang Chen, Tianle Cai, Tianhao Wu, Ruiqi Gao, Liwei Wang, Jason D. Lee
NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results show that our zero-shot random tickets outperform or attain a similar performance compared to existing initial tickets. We conduct experiments on three recently proposed pruning methods that can be classified as initial tickets. |
| Researcher Affiliation | Academia | Jingtong Su1, Yihang Chen2, Tianle Cai3,4, Tianhao Wu2 Ruiqi Gao3,4 Liwei Wang5,6, Jason D. Lee3, 1Yuanpei College, Peking University 2School of Mathematical Sciences, Peking University 3Department of Electrical Engineering, Princeton University 4Zhongguancun Haihua Institute for Frontier Information Technology 5Key Laboratory of Machine Perception, MOE, School of EECS, Peking University 6Center for Data Science, Peking University |
| Pseudocode | No | The paper does not contain structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | Our code is publicly available at https://github.com/Jingtong Su/sanity-checking-pruning. |
| Open Datasets | Yes | We conduct experiments on three recently proposed pruning methods that can be classified as initial tickets . We follow a standard training procedure as [23, 41, 17] throughout the entire paper on Res Net [19] and VGG [38]. The detailed setting can be found in Appendix B. ... on the original CIFAR-10 dataset. |
| Dataset Splits | No | The paper mentions 'test set' and 'standard training procedure' but does not provide specific details on training/validation/test dataset splits (e.g., percentages, sample counts, or explicit mention of a validation set) needed for reproduction. |
| 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 the use of general frameworks or models (e.g., Res Net, VGG) and specific pruning methods by citation, but it does not list any specific software packages or libraries with version numbers (e.g., Python, PyTorch, CUDA versions) needed to replicate the experiments. |
| Experiment Setup | Yes | The detailed setting can be found in Appendix B. |