{"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/pytorch-distributed-experiences-on","title":"PyTorch Distributed: Experiences on Accelerating Data Parallel Training","arxiv_id":"2006.15704","date":"2020-06-28","proceeding":null,"authors":["Shen Li","Yanli Zhao","Rohan Varma","Omkar Salpekar","Pieter Noordhuis","Teng Li","Adam Paszke","Jeff Smith","Brian Vaughan","Pritam Damania","Soumith Chintala"],"abstract":"This paper presents the design, implementation, and evaluation of the PyTorch distributed data parallel module. PyTorch is a widely-adopted scientific computing package used in deep learning research and applications. Recent advances in deep learning argue for the value of large datasets and large models, which necessitates the ability to scale out model training to more computational resources. Data parallelism has emerged as a popular solution for distributed training thanks to its straightforward principle and broad applicability. In general, the technique of distributed data parallelism replicates the model on every computational resource to generate gradients independently and then communicates those gradients at each iteration to keep model replicas consistent. Despite the conceptual simplicity of the technique, the subtle dependencies between computation and communication make it non-trivial to optimize the distributed training efficiency. As of v1.5, PyTorch natively provides several techniques to accelerate distributed data parallel, including bucketing gradients, overlapping computation with communication, and skipping gradient synchronization. Evaluations show that, when configured appropriately, the PyTorch distributed data parallel module attains near-linear scalability using 256 GPUs.","url_abs":"https://arxiv.org/abs/2006.15704v1","url_pdf":"https://arxiv.org/pdf/2006.15704v1.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":"pytorch-distributed-experiences-on","repo_url":"https://github.com/pytorch/pytorch/blob/master/torch/nn/parallel/distributed.py","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"jax","reach":null},{"paper_slug":"pytorch-distributed-experiences-on","repo_url":"https://github.com/ncfrey/litmatter","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"pytorch-distributed-experiences-on","repo_url":"https://github.com/lancelee82/necklace","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[{"method_slug":"pytorch-ddp","method_name":"PyTorch DDP"}],"datasets_introduced":[],"methods_introduced":[{"slug":"pytorch-ddp","name":"PyTorch DDP","full_name":"PyTorch DDP"}],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2006.15704","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2006.15704"}},"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/ncfrey/litmatter","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/lancelee82/necklace","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/pytorch/pytorch/blob/master/torch/nn/parallel/distributed.py","reach":null}],"summary":{"unverified":12},"by_repo_kind":{"listed":{"samples":9,"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":3,"samples":[{"code_sha256_prefix":"d8a9a401597da397","entry":"build_pg_cfg","repo":"lancelee82/necklace","repo_kind":"listed","path":"necklace/cuda/ncclgrp.py","file_url":"https://github.com/lancelee82/necklace/blob/HEAD/necklace/cuda/ncclgrp.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":"d8a9a401597da397"}},{"code_sha256_prefix":"d50063f8f0205b33","entry":"cuda_devarr_c_void_p","repo":"lancelee82/necklace","repo_kind":"listed","path":"necklace/cuda/nbutils.py","file_url":"https://github.com/lancelee82/necklace/blob/HEAD/necklace/cuda/nbutils.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":"d50063f8f0205b33"}},{"code_sha256_prefix":"3c44fe34338a17fa","entry":"cuda_to_device","repo":"lancelee82/necklace","repo_kind":"listed","path":"necklace/cuda/nbutils.py","file_url":"https://github.com/lancelee82/necklace/blob/HEAD/necklace/cuda/nbutils.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":"3c44fe34338a17fa"}},{"code_sha256_prefix":"08ce48d5e737dfa1","entry":"make_pg_by_nc","repo":"lancelee82/necklace","repo_kind":"listed","path":"necklace/cuda/ncclgrp.py","file_url":"https://github.com/lancelee82/necklace/blob/HEAD/necklace/cuda/ncclgrp.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":"08ce48d5e737dfa1"}},{"code_sha256_prefix":"c4e23fd5a2524f0e","entry":"make_pg_by_nk","repo":"lancelee82/necklace","repo_kind":"listed","path":"necklace/cuda/ncclgrp.py","file_url":"https://github.com/lancelee82/necklace/blob/HEAD/necklace/cuda/ncclgrp.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":"c4e23fd5a2524f0e"}},{"code_sha256_prefix":"91f1d429fa12a8c1","entry":"nccl_out_all_gather","repo":"lancelee82/necklace","repo_kind":"listed","path":"necklace/cuda/ncclutils.py","file_url":"https://github.com/lancelee82/necklace/blob/HEAD/necklace/cuda/ncclutils.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":"91f1d429fa12a8c1"}},{"code_sha256_prefix":"23734df31fcbe461","entry":"nk_comm_init_rank","repo":"lancelee82/necklace","repo_kind":"listed","path":"necklace/cuda/ncclfns.py","file_url":"https://github.com/lancelee82/necklace/blob/HEAD/necklace/cuda/ncclfns.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":"23734df31fcbe461"}},{"code_sha256_prefix":"358905d8834f7cfa","entry":"register_module_buffer_registration_hook","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"358905d8834f7cfa"}},{"code_sha256_prefix":"2e8b412bf625c913","entry":"register_module_module_registration_hook","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"2e8b412bf625c913"}},{"code_sha256_prefix":"0a1bb6fa560c4b8d","entry":"register_module_parameter_registration_hook","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"0a1bb6fa560c4b8d"}},{"code_sha256_prefix":"d00d171c543f55b4","entry":"seg_cb_nccl_ar","repo":"lancelee82/necklace","repo_kind":"listed","path":"necklace/cuda/ncclutils.py","file_url":"https://github.com/lancelee82/necklace/blob/HEAD/necklace/cuda/ncclutils.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":"d00d171c543f55b4"}},{"code_sha256_prefix":"c65797c72e9f5d34","entry":"tensor_data_ptr","repo":"lancelee82/necklace","repo_kind":"listed","path":"necklace/cuda/ncclutils.py","file_url":"https://github.com/lancelee82/necklace/blob/HEAD/necklace/cuda/ncclutils.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":"c65797c72e9f5d34"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}