SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2312.03526 · 2023

On the Diversity and Realism of Distilled Dataset: An Efficient Dataset Distillation Paradigm

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 10 functions out of this paper's own repositories and ran 3 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
lins-lab/rded canonical 1 of 6
shaoshitong/EDC reimplementation 2 of 3
copy not recorded — 0 of 1
FunctionStatusWhere it lives
accuracy Ran lins-lab/rded/validation/utils.py
code served (permissive licence) · get_code("eab4f74c38cc3c8f")
gkern Ran shaoshitong/EDC/Branch_ImageNet_10/recover/recover.py
pointer only (licence: NONE) · get_code("c0e0d0685f8dc4c7")
shift_list Ran shaoshitong/EDC/Branch_ImageNet_10/recover/recover.py
pointer only (licence: NONE) · get_code("871f8dea8a811189")
batched_forward Not yet run lins-lab/rded/synthesize/utils.py
code served (permissive licence) · get_code("b615cefacfc04d10")
epoch Not yet run shaoshitong/EDC/Branch_CIFAR_10/recover/baseline.py
pointer only (licence: NONE) · get_code("cf303e03d7c1efe6")
get_parameters Not yet run lins-lab/rded/validation/utils.py
code served (permissive licence) · get_code("9a700a3b149f4952")
get_state_dict Not yet run lins-lab/rded/synthesize/utils.py
code served (permissive licence) · get_code("b76a08bec5e010fd")
keep_top_k Not yet run lins-lab/rded/validation/utils.py
code served (permissive licence) · get_code("c209acc6c3e5877a")
pad Not yet run lins-lab/rded/synthesize/utils.py
code served (permissive licence) · get_code("680f3862844d8a00")
validate Not yet run this paper's copy was not recorded; identical code first harvested from VILA-Lab/SRe2L
pointer only · get_code("230fb2e9963063c9")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Contemporary machine learning requires training large neural networks on massive datasets and thus faces the challenges of high computational demands. Dataset distillation, as a recent emerging strategy, aims to compress real-world datasets for efficient training. However, this line of research currently struggle with large-scale and high-resolution datasets, hindering its practicality and feasibility. To this end, we re-examine the existing dataset distillation methods and identify three properties required for large-scale real-world applications, namely, realism, diversity, and efficiency. As a remedy, we propose RDED, a novel computationally-efficient yet effective data distillation paradigm, to enable both diversity and realism of the distilled data. Extensive empirical results over various neural architectures and datasets demonstrate the advancement of RDED: we can distill the full ImageNet-1K to a small dataset comprising 10 images per class within 7 minutes, achieving a notable 42% top-1 accuracy with ResNet-18 on a single RTX-4090 GPU (while the SOTA only achieves 21% but requires 6 hours).

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2312.03526")
get_code_for_paper("2312.03526")
have("2312.03526")

Connect an agent — have() is free.