{"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/revisiting-label-smoothing-and-knowledge","title":"Revisiting Label Smoothing and Knowledge Distillation Compatibility: What was Missing?","arxiv_id":"2206.14532","date":"2022-06-29","proceeding":null,"authors":["Keshigeyan Chandrasegaran","Ngoc-Trung Tran","Yunqing Zhao","Ngai-Man Cheung"],"abstract":"This work investigates the compatibility between label smoothing (LS) and knowledge distillation (KD). Contemporary findings addressing this thesis statement take dichotomous standpoints: Muller et al. (2019) and Shen et al. (2021b). Critically, there is no effort to understand and resolve these contradictory findings, leaving the primal question -- to smooth or not to smooth a teacher network? -- unanswered. The main contributions of our work are the discovery, analysis and validation of systematic diffusion as the missing concept which is instrumental in understanding and resolving these contradictory findings. This systematic diffusion essentially curtails the benefits of distilling from an LS-trained teacher, thereby rendering KD at increased temperatures ineffective. Our discovery is comprehensively supported by large-scale experiments, analyses and case studies including image classification, neural machine translation and compact student distillation tasks spanning across multiple datasets and teacher-student architectures. Based on our analysis, we suggest practitioners to use an LS-trained teacher with a low-temperature transfer to achieve high performance students. Code and models are available at https://keshik6.github.io/revisiting-ls-kd-compatibility/","url_abs":"https://arxiv.org/abs/2206.14532v1","url_pdf":"https://arxiv.org/pdf/2206.14532v1.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":"revisiting-label-smoothing-and-knowledge","repo_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"knowledge-distillation","task_name":"Knowledge Distillation"},{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"diffusion","method_name":"Diffusion"},{"method_slug":"knowledge-distillation","method_name":"Knowledge Distillation"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2206.14532","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2206.14532"}},"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":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_fixture":1,"ran":1,"unverified":14},"by_repo_kind":{"official":{"samples":16,"ran":2,"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":"131a82fd65128218","entry":"accuracy","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/imagenet/official_evaluation.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/imagenet/official_evaluation.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"131a82fd65128218"}},{"code_sha256_prefix":"6b5d363ab4dbf012","entry":"to_python_float","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/imagenet/train_student_dali.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/imagenet/train_student_dali.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6b5d363ab4dbf012"}},{"code_sha256_prefix":"4404239a85829756","entry":"adjust_learning_rate","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/cub/train_student.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/cub/train_student.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":"4404239a85829756"}},{"code_sha256_prefix":"e93943edd5a8472f","entry":"adjust_learning_rate","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/cub/train_teacher.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/cub/train_teacher.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":"e93943edd5a8472f"}},{"code_sha256_prefix":"fa122dd6d0671258","entry":"adjust_learning_rate","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/imagenet/official_evaluation.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/imagenet/official_evaluation.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":"fa122dd6d0671258"}},{"code_sha256_prefix":"205d6e975fc4fca7","entry":"get_cub_trained_teacher_for_cub_student","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/cub/utils.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/cub/utils.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":"205d6e975fc4fca7"}},{"code_sha256_prefix":"35f1a5ee27e2d2a0","entry":"get_imagenet_trained_teacher_for_cub_teacher","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/cub/utils.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/cub/utils.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":"35f1a5ee27e2d2a0"}},{"code_sha256_prefix":"fbaa9853176fc776","entry":"get_imagenet_trained_teacher_for_imagenet_student","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/imagenet/utils.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/imagenet/utils.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":"fbaa9853176fc776"}},{"code_sha256_prefix":"06ec1746d47a9901","entry":"get_penultimate_fts","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/visualization/alpha-LS-KD_imagenet_centroids.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/visualization/alpha-LS-KD_imagenet_centroids.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":"06ec1746d47a9901"}},{"code_sha256_prefix":"afe574de7fe140de","entry":"get_test_loader","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/visualization/data_loader_imagenet.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/visualization/data_loader_imagenet.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":"afe574de7fe140de"}},{"code_sha256_prefix":"33ec9768e54f5c4a","entry":"load_model_kd","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/visualization/alpha-LS-KD_imagenet_centroids.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/visualization/alpha-LS-KD_imagenet_centroids.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":"33ec9768e54f5c4a"}},{"code_sha256_prefix":"1517519d4d7d27cf","entry":"load_model_ls","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/visualization/alpha-LS-KD_imagenet_centroids.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/visualization/alpha-LS-KD_imagenet_centroids.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":"1517519d4d7d27cf"}},{"code_sha256_prefix":"2a0d7e917240db2a","entry":"train","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/cub/train_student.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/cub/train_student.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":"2a0d7e917240db2a"}},{"code_sha256_prefix":"020b5acc2cea526a","entry":"validate","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/cub/train_student.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/cub/train_student.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":"020b5acc2cea526a"}},{"code_sha256_prefix":"889998947e765d45","entry":"validate","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/imagenet/official_evaluation.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/imagenet/official_evaluation.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":"889998947e765d45"}},{"code_sha256_prefix":"00c8fdc297e7e272","entry":"validate","repo":"sutd-visual-computing-group/LS-KD-compatibility","repo_kind":"official","path":"src/image_classification/imagenet/.ipynb_checkpoints/official_evaluation-checkpoint.py","file_url":"https://github.com/sutd-visual-computing-group/LS-KD-compatibility/blob/HEAD/src/image_classification/imagenet/.ipynb_checkpoints/official_evaluation-checkpoint.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":"00c8fdc297e7e272"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}