{"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/fine-tuning-language-models-over-slow","title":"Fine-tuning Language Models over Slow Networks using Activation Compression with Guarantees","arxiv_id":"2206.01299","date":"2022-06-02","proceeding":null,"authors":["Jue Wang","Binhang Yuan","Luka Rimanic","Yongjun He","Tri Dao","Beidi Chen","Christopher Re","Ce Zhang"],"abstract":"Communication compression is a crucial technique for modern distributed learning systems to alleviate their communication bottlenecks over slower networks. Despite recent intensive studies of gradient compression for data parallel-style training, compressing the activations for models trained with pipeline parallelism is still an open problem. In this paper, we propose AC-SGD, a novel activation compression algorithm for communication-efficient pipeline parallelism training over slow networks. Different from previous efforts in activation compression, instead of compressing activation values directly, AC-SGD compresses the changes of the activations. This allows us to show, to the best of our knowledge for the first time, that one can still achieve $O(1/\\sqrt{T})$ convergence rate for non-convex objectives under activation compression, without making assumptions on gradient unbiasedness that do not hold for deep learning models with non-linear activation functions.We then show that AC-SGD can be optimized and implemented efficiently, without additional end-to-end runtime overhead.We evaluated AC-SGD to fine-tune language models with up to 1.5 billion parameters, compressing activations to 2-4 bits.AC-SGD provides up to 4.3X end-to-end speed-up in slower networks, without sacrificing model quality. Moreover, we also show that AC-SGD can be combined with state-of-the-art gradient compression algorithms to enable \"end-to-end communication compression: All communications between machines, including model gradients, forward activations, and backward gradients are compressed into lower precision.This provides up to 4.9X end-to-end speed-up, without sacrificing model quality.","url_abs":"https://arxiv.org/abs/2206.01299v3","url_pdf":"https://arxiv.org/pdf/2206.01299v3.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":"fine-tuning-language-models-over-slow","repo_url":"https://github.com/DS3Lab/AC-SGD","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2206.01299","atlas_url":"https://app.syntology.ai/?focus=2206.01299","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2206.01299"}},"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/DS3Lab/AC-SGD","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":3,"unverified":2},"by_repo_kind":{"official":{"samples":5,"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":"0728103d07b148cd","entry":"build_relative_position","repo":"DS3Lab/AC-SGD","repo_kind":"official","path":"modules/deberta_modules.py","file_url":"https://github.com/DS3Lab/AC-SGD/blob/HEAD/modules/deberta_modules.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":"0728103d07b148cd"}},{"code_sha256_prefix":"b9f88ec96172331a","entry":"gpt_loss_func","repo":"DS3Lab/AC-SGD","repo_kind":"official","path":"modules/gpt_modules.py","file_url":"https://github.com/DS3Lab/AC-SGD/blob/HEAD/modules/gpt_modules.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":"b9f88ec96172331a"}},{"code_sha256_prefix":"3d0181ea3d80faa8","entry":"make_log_bucket_position","repo":"DS3Lab/AC-SGD","repo_kind":"official","path":"modules/deberta_modules.py","file_url":"https://github.com/DS3Lab/AC-SGD/blob/HEAD/modules/deberta_modules.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":"3d0181ea3d80faa8"}},{"code_sha256_prefix":"6f04b420da6e0e2a","entry":"build_deberta_tokenizer","repo":"DS3Lab/AC-SGD","repo_kind":"official","path":"modules/tokenizer.py","file_url":"https://github.com/DS3Lab/AC-SGD/blob/HEAD/modules/tokenizer.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":"6f04b420da6e0e2a"}},{"code_sha256_prefix":"b0c8736d6680465b","entry":"build_tokenizer","repo":"DS3Lab/AC-SGD","repo_kind":"official","path":"modules/tokenizer.py","file_url":"https://github.com/DS3Lab/AC-SGD/blob/HEAD/modules/tokenizer.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":"b0c8736d6680465b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}