{"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/training-wide-residual-networks-for","title":"Training wide residual networks for deployment using a single bit for each weight","arxiv_id":"1802.08530","date":"2018-02-23","proceeding":"ICLR 2018 1","authors":["Mark D. McDonnell"],"abstract":"For fast and energy-efficient deployment of trained deep neural networks on\nresource-constrained embedded hardware, each learned weight parameter should\nideally be represented and stored using a single bit. Error-rates usually\nincrease when this requirement is imposed. Here, we report large improvements\nin error rates on multiple datasets, for deep convolutional neural networks\ndeployed with 1-bit-per-weight. Using wide residual networks as our main\nbaseline, our approach simplifies existing methods that binarize weights by\napplying the sign function in training; we apply scaling factors for each layer\nwith constant unlearned values equal to the layer-specific standard deviations\nused for initialization. For CIFAR-10, CIFAR-100 and ImageNet, and models with\n1-bit-per-weight requiring less than 10 MB of parameter memory, we achieve\nerror rates of 3.9%, 18.5% and 26.0% / 8.5% (Top-1 / Top-5) respectively. We\nalso considered MNIST, SVHN and ImageNet32, achieving 1-bit-per-weight test\nresults of 0.27%, 1.9%, and 41.3% / 19.1% respectively. For CIFAR, our error\nrates halve previously reported values, and are within about 1% of our\nerror-rates for the same network with full-precision weights. For networks that\noverfit, we also show significant improvements in error rate by not learning\nbatch normalization scale and offset parameters. This applies to both full\nprecision and 1-bit-per-weight networks. Using a warm-restart learning-rate\nschedule, we found that training for 1-bit-per-weight is just as fast as\nfull-precision networks, with better accuracy than standard schedules, and\nachieved about 98%-99% of peak performance in just 62 training epochs for\nCIFAR-10/100. For full training code and trained models in MATLAB, Keras and\nPyTorch see https://github.com/McDonnell-Lab/1-bit-per-weight/ .","url_abs":"http://arxiv.org/abs/1802.08530v1","url_pdf":"http://arxiv.org/pdf/1802.08530v1.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":"training-wide-residual-networks-for","repo_url":"https://github.com/McDonnell-Lab/1-bit-per-weight","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"training-wide-residual-networks-for","repo_url":"https://github.com/osmr/imgclsmob","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mxnet","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"training-wide-residual-networks-for","repo_url":"https://github.com/szagoruyko/binary-wide-resnet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"training-wide-residual-networks-for","repo_url":"https://github.com/Mind23-2/MindCode-4/tree/main/wide_resnet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"training-wide-residual-networks-for","repo_url":"https://github.com/code-implementation1/Code9/tree/main/wide_resnet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1802.08530","atlas_url":"https://app.syntology.ai/?focus=1802.08530","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1802.08530"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/szagoruyko/binary-wide-resnet","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/osmr/imgclsmob","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/code-implementation1/Code9/tree/main/wide_resnet","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Mind23-2/MindCode-4/tree/main/wide_resnet","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/McDonnell-Lab/1-bit-per-weight","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":5},"by_repo_kind":{"official":{"samples":5,"ran":0,"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":"d8bcd6f65945b68b","entry":"GetDataGen","repo":"McDonnell-Lab/1-bit-per-weight","repo_kind":"official","path":"Utils.py","file_url":"https://github.com/McDonnell-Lab/1-bit-per-weight/blob/HEAD/Utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d8bcd6f65945b68b"}},{"code_sha256_prefix":"f9677443f26ffc60","entry":"cutout","repo":"McDonnell-Lab/1-bit-per-weight","repo_kind":"official","path":"Utils.py","file_url":"https://github.com/McDonnell-Lab/1-bit-per-weight/blob/HEAD/Utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f9677443f26ffc60"}},{"code_sha256_prefix":"140e7ae9101fa563","entry":"resnet","repo":"McDonnell-Lab/1-bit-per-weight","repo_kind":"official","path":"ResNetModel.py","file_url":"https://github.com/McDonnell-Lab/1-bit-per-weight/blob/HEAD/ResNetModel.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"140e7ae9101fa563"}},{"code_sha256_prefix":"433abcfe62c095dc","entry":"resnet_layer","repo":"McDonnell-Lab/1-bit-per-weight","repo_kind":"official","path":"ResNetModel.py","file_url":"https://github.com/McDonnell-Lab/1-bit-per-weight/blob/HEAD/ResNetModel.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"433abcfe62c095dc"}},{"code_sha256_prefix":"927619a32a291ba3","entry":"resnet_layer_srelu","repo":"McDonnell-Lab/1-bit-per-weight","repo_kind":"official","path":"ResNetModel.py","file_url":"https://github.com/McDonnell-Lab/1-bit-per-weight/blob/HEAD/ResNetModel.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"927619a32a291ba3"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}