{"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/moonshine-distilling-with-cheap-convolutions","title":"Moonshine: Distilling with Cheap Convolutions","arxiv_id":"1711.02613","date":"2017-11-07","proceeding":"NeurIPS 2018 12","authors":["Elliot J. Crowley","Gavin Gray","Amos Storkey"],"abstract":"Many engineers wish to deploy modern neural networks in memory-limited\nsettings; but the development of flexible methods for reducing memory use is in\nits infancy, and there is little knowledge of the resulting cost-benefit. We\npropose structural model distillation for memory reduction using a strategy\nthat produces a student architecture that is a simple transformation of the\nteacher architecture: no redesign is needed, and the same hyperparameters can\nbe used. Using attention transfer, we provide Pareto curves/tables for\ndistillation of residual networks with four benchmark datasets, indicating the\nmemory versus accuracy payoff. We show that substantial memory savings are\npossible with very little loss of accuracy, and confirm that distillation\nprovides student network performance that is better than training that student\narchitecture directly on data.","url_abs":"http://arxiv.org/abs/1711.02613v4","url_pdf":"http://arxiv.org/pdf/1711.02613v4.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":"moonshine-distilling-with-cheap-convolutions","repo_url":"https://github.com/BayesWatch/pytorch-moonshine","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1711.02613","atlas_url":"https://app.syntology.ai/?focus=1711.02613","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1711.02613"}},"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/BayesWatch/pytorch-moonshine","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_honours":2,"ran_violates":1,"unverified":6},"by_repo_kind":{"official":{"samples":9,"ran":3,"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":"7556a5bddc992e63","entry":"at","repo":"BayesWatch/pytorch-moonshine","repo_kind":"official","path":"funcs.py","file_url":"https://github.com/BayesWatch/pytorch-moonshine/blob/HEAD/funcs.py","link_basis":"plan_row","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7556a5bddc992e63"}},{"code_sha256_prefix":"9620e20bbcbbb1d0","entry":"at_loss","repo":"BayesWatch/pytorch-moonshine","repo_kind":"official","path":"funcs.py","file_url":"https://github.com/BayesWatch/pytorch-moonshine/blob/HEAD/funcs.py","link_basis":"plan_row","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9620e20bbcbbb1d0"}},{"code_sha256_prefix":"f6484efda3ef4ceb","entry":"distillation","repo":"BayesWatch/pytorch-moonshine","repo_kind":"official","path":"funcs.py","file_url":"https://github.com/BayesWatch/pytorch-moonshine/blob/HEAD/funcs.py","link_basis":"plan_row","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f6484efda3ef4ceb"}},{"code_sha256_prefix":"5115f85144261493","entry":"ACDC","repo":"BayesWatch/pytorch-moonshine","repo_kind":"official","path":"models/blocks.py","file_url":"https://github.com/BayesWatch/pytorch-moonshine/blob/HEAD/models/blocks.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":"5115f85144261493"}},{"code_sha256_prefix":"dfeb82e670f646bf","entry":"conv_function","repo":"BayesWatch/pytorch-moonshine","repo_kind":"official","path":"models/blocks.py","file_url":"https://github.com/BayesWatch/pytorch-moonshine/blob/HEAD/models/blocks.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":"dfeb82e670f646bf"}},{"code_sha256_prefix":"76cc74efee42ba21","entry":"get_num_gen","repo":"BayesWatch/pytorch-moonshine","repo_kind":"official","path":"count_flops.py","file_url":"https://github.com/BayesWatch/pytorch-moonshine/blob/HEAD/count_flops.py","link_basis":"harvester_set","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":"76cc74efee42ba21"}},{"code_sha256_prefix":"26ff085b343fa39e","entry":"is_leaf","repo":"BayesWatch/pytorch-moonshine","repo_kind":"official","path":"count_flops.py","file_url":"https://github.com/BayesWatch/pytorch-moonshine/blob/HEAD/count_flops.py","link_basis":"harvester_set","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":"26ff085b343fa39e"}},{"code_sha256_prefix":"5c4dab079bf6276e","entry":"is_pruned","repo":"BayesWatch/pytorch-moonshine","repo_kind":"official","path":"count_flops.py","file_url":"https://github.com/BayesWatch/pytorch-moonshine/blob/HEAD/count_flops.py","link_basis":"harvester_set","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":"5c4dab079bf6276e"}},{"code_sha256_prefix":"88983196c12d6ae2","entry":"parse_options","repo":"BayesWatch/pytorch-moonshine","repo_kind":"official","path":"models/wide_resnet.py","file_url":"https://github.com/BayesWatch/pytorch-moonshine/blob/HEAD/models/wide_resnet.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":"88983196c12d6ae2"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}