{ "profile": "wrs-published-research-view-1", "edition": "ratings-e039106f30dec177", "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS", "entry": { "id": "pangu-alpha-team-autoparallel-chinese-system-2021", "title": "PanGu-Alpha Team: five-axis auto-parallel Chinese language-model training", "summary": "Integrated data, operator-model, pipeline and optimizer parallelism with rematerialization in MindSpore Auto-parallel for large Chinese autoregressive model training on Ascend accelerators.", "boundary": "Transformer/GPT pretraining, parallelization and rematerialization are inherited. Acquired delta is integrated training system and Chinese model resource, not each parallel dimension. Processor/model scales and language evaluations are historical reported experiments, not current performance guarantees.", "area": "Distributed foundation-model training systems", "topic": "Artificial intelligence", "year": 2021, "lane": "HUMAN", "result": "EMPIRICAL", "package_id": "pangu-alpha-team-autoparallel-chinese-system-2021", "target": { "record_id": "urn:uuid:707c3b3e-5513-4bf6-a531-bdfb2a4d19a7", "version": 1, "content_hash": "sha256:c40077ed8805d074df55e7416d8fe32ef70f0a28f502f9f878f7be865301b4b3" }, "humans": [ "Wei Zeng", "Xiaozhe Ren", "Teng Su", "Hui Wang", "Yi Liao", "Zhiwei Wang", "Xin Jiang", "Zhenzhang Yang", "Kaisheng Wang", "Xiaoda Zhang", "Chen Li", "Ziyan Gong", "Yifan Yao", "Xinjing Huang", "Jun Wang", "Jianfeng Yu", "Qi Guo", "Yue Yu", "Yan Zhang", "Jin Wang", "Hengtao Tao", "Dasen Yan", "Zexuan Yi", "Fang Peng", "Fangqing Jiang", "Han Zhang", "Lingfeng Deng", "Yehong Zhang", "Zhe Lin", "Chao Zhang", "Shaojie Zhang", "Mingyue Guo", "Shanzhi Gu", "Gaojun Fan", "Yaowei Wang", "Xuefeng Jin", "Qun Liu", "Yonghong Tian" ], "authors_complete": true, "human_role_note": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.", "agents": [], "related": [], "subresults": [], "scientific_assessment": { "source_review": "SOURCE_CURATED", "scientific_review": "NOT_REVIEWED_BY_WRS", "formal_replay": "NOT_RUN", "provider_attestation": "NONE" }, "sources": [ { "id": "cur20261015-arxiv-2104-12369v1", "title": "PanGu-$α$: Large-scale Autoregressive Pretrained Chinese Language Models with Auto-parallel Computation", "url": "https://arxiv.org/abs/2104.12369v1" } ], "points": "115000", "published_grade": { "id": "pangu-alpha-team-autoparallel-chinese-system-2021", "entry_id": "pangu-alpha-team-autoparallel-chinese-system-2021", "band": "A", "modifier": null, "target": { "record_id": "urn:uuid:707c3b3e-5513-4bf6-a531-bdfb2a4d19a7", "version": 1, "content_hash": "sha256:c40077ed8805d074df55e7416d8fe32ef70f0a28f502f9f878f7be865301b4b3" }, "what_changed": "Integrated data, operator-model, pipeline and optimizer parallelism with rematerialization in MindSpore Auto-parallel for large Chinese autoregressive model training on Ascend accelerators.", "scope": "Integrated data, operator-model, pipeline and optimizer parallelism with rematerialization in MindSpore Auto-parallel for large Chinese autoregressive model training on Ascend accelerators. Transformer/GPT pretraining, parallelization and rematerialization are inherited. Acquired delta is integrated training system and Chinese model resource, not each parallel dimension. Processor/model scales and language evaluations are historical reported experiments, not current performance guarantees.", "status": "PROVISIONAL_EDITORIAL", "scientific_correctness_certified": false }, "assessment_disposition": { "status": "ASSESSED", "label": "Provisionally assessed", "reason": "A current grade and separate reference score identify this exact contribution.", "target": { "record_id": "urn:uuid:707c3b3e-5513-4bf6-a531-bdfb2a4d19a7", "version": 1, "content_hash": "sha256:c40077ed8805d074df55e7416d8fe32ef70f0a28f502f9f878f7be865301b4b3" }, "edition": "ratings-e039106f30dec177" } }, "version": { "kind": "PUBLIC_RECORD_VIEW", "previous": null, "published_at": "2026-10-16T00:00:00Z", "publisher": { "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33", "role": "WRS_EDITORIAL_PUBLICATION" }, "record": { "contributors": [ { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "organization", "name": "PanGu-α Team", "note": "Cited collaboration; the full byline is retained in the structured atlas and imported ledger, committed by hash in the narrative.", "roles": [ "discoverer" ], "source": { "citation": "PanGu-$α$: Large-scale Autoregressive Pretrained Chinese Language Models with Auto-parallel Computation", "url": "https://arxiv.org/abs/2104.12369v1" } } ], "evidence": [ { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-0", "kind": "source_reference", "source": { "citation": "PanGu-$α$: Large-scale Autoregressive Pretrained Chinese Language Models with Auto-parallel Computation", "url": "https://arxiv.org/abs/2104.12369v1" } } ], "origin": "HUMAN_LEGACY", "record_profile": "wrs-local-research-notes-0.1", "record_type": "research_note", "rrp_version": "0.1", "scope": { "description": "Transformer/GPT pretraining, parallelization and rematerialization are inherited. Acquired delta is integrated training system and Chinese model resource, not each parallel dimension. Processor/model scales and language evaluations are historical reported experiments, not current performance guarantees.", "domain": "Distributed foundation-model training systems" }, "sources": [ { "citation": "PanGu-$α$: Large-scale Autoregressive Pretrained Chinese Language Models with Auto-parallel Computation", "url": "https://arxiv.org/abs/2104.12369v1" } ], "statement": "# PanGu-Alpha Team: five-axis auto-parallel Chinese language-model training\n\nIntegrated data, operator-model, pipeline and optimizer parallelism with rematerialization in MindSpore Auto-parallel for large Chinese autoregressive model training on Ascend accelerators.\n\n## Scientific scope\n\nTransformer/GPT pretraining, parallelization and rematerialization are inherited. Acquired delta is integrated training system and Chinese model resource, not each parallel dimension. Processor/model scales and language evaluations are historical reported experiments, not current performance guarantees.\n\n## Reading status\n\nThis is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.\n\n## People and systems\n\nWei Zeng; Xiaozhe Ren; Teng Su; Hui Wang; Yi Liao; Zhiwei Wang; Xin Jiang; Zhenzhang Yang; Kaisheng Wang; Xiaoda Zhang; Chen Li; Ziyan Gong; Yifan Yao; Xinjing Huang; Jun Wang; Jianfeng Yu; Qi Guo; Yue Yu; Yan Zhang; Jin Wang; Hengtao Tao; Dasen Yan; Zexuan Yi; Fang Peng; Fangqing Jiang; Han Zhang; Lingfeng Deng; Yehong Zhang; Zhe Lin; Chao Zhang; Shaojie Zhang; Mingyue Guo; Shanzhi Gu; Gaojun Fan; Yaowei Wang; Xuefeng Jin; Qun Liu; Yonghong Tian\n\nSource-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.\n\n\n\n## Original sources\n\n- [PanGu-$α$: Large-scale Autoregressive Pretrained Chinese Language Models with Auto-parallel Computation](https://arxiv.org/abs/2104.12369v1)\n", "summary": "Integrated data, operator-model, pipeline and optimizer parallelism with rematerialization in MindSpore Auto-parallel for large Chinese autoregressive model training on Ascend accelerators.", "title": "PanGu-Alpha Team: five-axis auto-parallel Chinese language-model training" }, "record_id": "urn:uuid:707c3b3e-5513-4bf6-a531-bdfb2a4d19a7", "version": 1, "source_ref": { "record_id": "urn:uuid:707c3b3e-5513-4bf6-a531-bdfb2a4d19a7", "version": 1, "content_hash": "sha256:c40077ed8805d074df55e7416d8fe32ef70f0a28f502f9f878f7be865301b4b3" }, "export_view": "wrs-public-editorial-view-1", "public_view_hash": "sha256:7424a87520b6c6ef7e3f73d9321809918044bc2bdd4c2b03cfe1ac7afc2b5dd4" }, "target": { "record_id": "urn:uuid:707c3b3e-5513-4bf6-a531-bdfb2a4d19a7", "version": 1, "content_hash": "sha256:c40077ed8805d074df55e7416d8fe32ef70f0a28f502f9f878f7be865301b4b3" }, "published_grade": { "profile": "wrs-published-grade-1", "edition": "2026-10-28.1", "id": "pangu-alpha-team-autoparallel-chinese-system-2021", "entry_id": "pangu-alpha-team-autoparallel-chinese-system-2021", "revision": 1, "title": "PanGu-Alpha Team: five-axis auto-parallel Chinese language-model training", "context": "Distributed foundation-model training systems", "track": "Human-designed research method or resource", "band": "A", "modifier": null, "status": "PROVISIONAL_EDITORIAL", "assessed_on": "2026-10-16", "evidence_cutoff": "2026-10-16", "achievement_state": "SUPPORTED_SCOPED_RESULT", "what_changed": "Integrated data, operator-model, pipeline and optimizer parallelism with rematerialization in MindSpore Auto-parallel for large Chinese autoregressive model training on Ascend accelerators.", "scope": "Integrated data, operator-model, pipeline and optimizer parallelism with rematerialization in MindSpore Auto-parallel for large Chinese autoregressive model training on Ascend accelerators. Transformer/GPT pretraining, parallelization and rematerialization are inherited. Acquired delta is integrated training system and Chinese model resource, not each parallel dimension. Processor/model scales and language evaluations are historical reported experiments, not current performance guarantees.", "evidence_summary": "The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Transformer/GPT pretraining, parallelization and rematerialization are inherited. Acquired delta is integrated training system and Chinese model resource, not each parallel dimension. Processor/model scales and language evaluations are historical reported experiments, not current performance guarantees.", "references": [ { "title": "PanGu-$α$: Large-scale Autoregressive Pretrained Chinese Language Models with Auto-parallel Computation", "url": "https://arxiv.org/abs/2104.12369v1", "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification." } ], "roles": [ { "who": "Source-listed authors and technical contributors", "credit": "SCIENCE", "did": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities." } ], "target": { "record_id": "urn:uuid:707c3b3e-5513-4bf6-a531-bdfb2a4d19a7", "version": 1, "content_hash": "sha256:c40077ed8805d074df55e7416d8fe32ef70f0a28f502f9f878f7be865301b4b3" }, "scientific_correctness_certified": false, "independent_expert_calibration": false, "method": "PROPRIETARY", "points": "115000" }, "canonical_agent_submission": false, "scientific_verification_awarded": false, "reader_guide": null, "reading_status": "SOURCE_LINKED_NOTE" }