{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/ac-dc-alternating-compressed-decompressed","title":"AC/DC: Alternating Compressed/DeCompressed Training of Deep Neural Networks","arxiv_id":"2106.12379","date":"2021-06-23","proceeding":"NeurIPS 2021 12","authors":["Alexandra Peste","Eugenia Iofinova","Adrian Vladu","Dan Alistarh"],"abstract":"The increasing computational requirements of deep neural networks (DNNs) have led to significant interest in obtaining DNN models that are sparse, yet accurate. Recent work has investigated the even harder case of sparse training, where the DNN weights are, for as much as possible, already sparse to reduce computational costs during training. Existing sparse training methods are often empirical and can have lower accuracy relative to the dense baseline. In this paper, we present a general approach called Alternating Compressed/DeCompressed (AC/DC) training of DNNs, demonstrate convergence for a variant of the algorithm, and show that AC/DC outperforms existing sparse training methods in accuracy at similar computational budgets; at high sparsity levels, AC/DC even outperforms existing methods that rely on accurate pre-trained dense models. An important property of AC/DC is that it allows co-training of dense and sparse models, yielding accurate sparse-dense model pairs at the end of the training process. This is useful in practice, where compressed variants may be desirable for deployment in resource-constrained settings without re-doing the entire training flow, and also provides us with insights into the accuracy gap between dense and compressed models. The code is available at: https://github.com/IST-DASLab/ACDC .","url_abs":"https://arxiv.org/abs/2106.12379v2","url_pdf":"https://arxiv.org/pdf/2106.12379v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"ac-dc-alternating-compressed-decompressed","repo_url":"https://github.com/IST-DASLab/ACDC","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"ac-dc-alternating-compressed-decompressed","repo_url":"https://github.com/IST-DASLab/sparseprop","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"network-pruning","task_name":"Network Pruning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/network-pruning-on-cifar-100","task":"Network Pruning","dataset":"CIFAR-100","model":"Dense","rank_in_archive_order":1,"of":5,"metrics":{"Accuracy":"79"},"uses_additional_data":false},{"leaderboard":"/sota/network-pruning-on-cifar-100","task":"Network Pruning","dataset":"CIFAR-100","model":"AC/DC","rank_in_archive_order":2,"of":5,"metrics":{"Accuracy":"78.2"},"uses_additional_data":false},{"leaderboard":"/sota/network-pruning-on-imagenet","task":"Network Pruning","dataset":"ImageNet","model":"ResNet50","rank_in_archive_order":14,"of":16,"metrics":{"Accuracy":"73.14"},"uses_additional_data":false},{"leaderboard":"/sota/network-pruning-on-imagenet-resnet-50-90","task":"Network Pruning","dataset":"ImageNet - ResNet 50 - 90% sparsity","model":"AC/DC","rank_in_archive_order":4,"of":9,"metrics":{"Top-1 Accuracy":"75.64"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2106.12379","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2106.12379"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"deterministic:regex_extraction","url":"https://github.com/IST-DASLab/ACDC","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/IST-DASLab/sparseprop","reach":null}],"summary":{"ran":2,"unverified":3},"by_repo_kind":{"official":{"samples":1,"ran":1,"repositories":1},"listed":{"samples":4,"ran":1,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"bd353d9bbcb08c1e","entry":"GradualNormPrunerSGD","repo":"IST-DASLab/ACDC","repo_kind":"official","path":"optimization/gradual_norm_reduction_pruner.py","file_url":"https://github.com/IST-DASLab/ACDC/blob/HEAD/optimization/gradual_norm_reduction_pruner.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"bd353d9bbcb08c1e"}},{"code_sha256_prefix":"1e358405d88a676c","entry":"SparseConvFunction","repo":"IST-DASLab/sparseprop","repo_kind":"listed","path":"sparseprop/modules/conv2d.py","file_url":"https://github.com/IST-DASLab/sparseprop/blob/HEAD/sparseprop/modules/conv2d.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"1e358405d88a676c"}},{"code_sha256_prefix":"0eb137fa4fe54737","entry":"SparseConv2d","repo":"IST-DASLab/sparseprop","repo_kind":"listed","path":"sparseprop/modules/conv2d.py","file_url":"https://github.com/IST-DASLab/sparseprop/blob/HEAD/sparseprop/modules/conv2d.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"0eb137fa4fe54737"}},{"code_sha256_prefix":"bc744953cf54aea4","entry":"from_sparse_format_conv2d","repo":"IST-DASLab/sparseprop","repo_kind":"listed","path":"sparseprop/modules/conv2d.py","file_url":"https://github.com/IST-DASLab/sparseprop/blob/HEAD/sparseprop/modules/conv2d.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"bc744953cf54aea4"}},{"code_sha256_prefix":"55366aa103be949f","entry":"to_sparse_format_conv2d","repo":"IST-DASLab/sparseprop","repo_kind":"listed","path":"sparseprop/modules/conv2d.py","file_url":"https://github.com/IST-DASLab/sparseprop/blob/HEAD/sparseprop/modules/conv2d.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"55366aa103be949f"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}