mirror of
https://github.com/harvard-edge/cs249r_book.git
synced 2026-05-08 02:28:25 -05:00
[GH-ISSUE #304] Suggestion: Cite the following papers #1351
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @FinAminToastCrunch on GitHub (Jul 2, 2024).
Original GitHub issue: https://github.com/harvard-edge/cs249r_book/issues/304
In reference to Ethical Considerations on Pruning, I would suggest citing:
Tran, Cuong, et al. "Pruning has a disparate impact on model accuracy." Advances in neural information processing systems 35 (2022): 17652-17664.
As it discusses how pruning can harm accuracy with respect to inferencing on certain ethnicities.
In reference to Establishing Criteria for Pruning, I believe these two papers would be relevant to cite for the portions on importance scores as they tie together more recent work concerning gradient flow and neuron magnitude:
J. Rachwan, D. Z ̈ugner, B. Charpentier, S. Geisler, M. Ayle, and
S. G ̈unnemann, “Winning the lottery ahead of time: Efficient early
network pruning,” in International Conference on Machine Learning,
ICML 2022, 17-23 July 2022, Baltimore, Maryland, USA (K. Chaudhuri,
S. Jegelka, L. Song, C. Szepesv ́ari, G. Niu, and S. Sabato, eds.),
vol. 162 of Proceedings of Machine Learning Research, pp. 18293–
18309, PMLR, 2022.
E. S. Lubana and R. P. Dick, “A gradient flow framework for analyzing
network pruning,” CoRR, vol. abs/2009.11839, 2020.
@profvjreddi commented on GitHub (Jul 4, 2024):
@FinAminToastCrunch The link you shared is not working. Could you update the link so I can take at which part you are referring to here.
@FinAminToastCrunch commented on GitHub (Jul 4, 2024):
Fixed:
https://harvard-edge.github.io/cs249r_book/contents/optimizations/optimizations.html#legal-and-ethical-considerations
@profvjreddi commented on GitHub (Jul 4, 2024):
Thanks Fin! Do you think you could issue a PR so that I could review it and
merge it in? I will edit it if needed after you issue a PR.
On Thu, Jul 04, 2024 at 12:32 PM, Fin Amin @.***> wrote:
@profvjreddi commented on GitHub (Jul 4, 2024):
And just as FYI. I typically avoid incorporating overly cutting-edge
material into the book. The criterion for inclusion is that the industry
must be actively adopting and utilizing the methods discussed. This
approach ensures the content remains relevant and avoids delving into an
excessive number of rapidly evolving topics. Does that make sense?
On Thu, Jul 04, 2024 at 12:44 PM, Janapa Reddi, Vijay @.***>
wrote:
@FinAminToastCrunch commented on GitHub (Jul 4, 2024):
Done!
And yes, makes sense.
@profvjreddi commented on GitHub (Jul 8, 2024):
Dealt with in issue https://github.com/harvard-edge/cs249r_book/pull/306 and PR
@profvjreddi commented on GitHub (Jul 30, 2024):
Thanks @FinAminToastCrunch I finally got a chance to review the changes and make some tweaks an close it.