{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:IAWCD73NKD3BCJFE6D3BSA4PYG","short_pith_number":"pith:IAWCD73N","schema_version":"1.0","canonical_sha256":"402c21ff6d50f61124a4f0f619038fc18f4f158289990eb722a071cce640a354","source":{"kind":"arxiv","id":"2607.20145","version":1},"attestation_state":"computed","paper":{"title":"SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Abdullin Amir, Baotian Hu, Barkova Maria, Bingwei Zhang, Chang Liu, Chenwei Wu, Cuihua Hu, Dongfang Li, Dong Zhang, Fangxue Zhang, Haizhou Li, Hongzhang Yang, Igor Vasilyev, Jiajun Liu, Jian Meng, Jie Yu, Jing Li, Jinhua Zhou, Jinmin Xiang, Junjie Huang, Lin Hu, Linxin Yang, Linyuan Qiu, Li Zeng, Mingzhe Yang, Min Liu, Min Zhang, Nosov Ivan, Qian Chen, Ruoyu Sun, Shihao Hong, Shihao Zeng, Shijun Qin, Sihang Chen, Tao Guo, Taolue Chen, Tian Ding, Tianxiang Fang, Tolstykh Vasiliy, Ushakov Anton, Weihua Dai, Weijian Sun, Weiyan Kong, Wenhan Luo, Wenlong Jin, Xianfei Jin, Xiaodong Luo, Xin Chen, Xin Zhang, Xuhui Chen, Xutong Zhao, Yangbo Guo, Yingmeng Gao, Yiting Wang, Yucheng Xie, Zhengjun Yue, Zhengxuan Lu, Zheng Zhang, Zhichen Zhou, Zhihang Lin, Zhiquan Luo, Zhixiong Ning, Zhuohan Wang, Zirong Zeng, Ziwei Zhu","submitted_at":"2026-07-22T13:49:17Z","abstract_excerpt":"Full-parameter post-training of trillion-parameter-scale MoE models introduces substantial system-level challenges for large-scale distributed training, including severe memory pressure, non-overlapped communication overhead, and inefficient kernel execution. While most large-scale LLM training systems are built around GPU-based clusters, this report presents an end-to-end optimization practice on the Ascend NPU SuperPOD. Using the DeepSeek-V4 model family as the target workload, we develop a hierarchical optimization framework spanning model-level parallelism, computation-communication orches"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2607.20145","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-07-22T13:49:17Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"4a3f079b62cdd836b3ef143bd48a32fb3a39e1d92c06ea4c99a77f0016ecf894","abstract_canon_sha256":"8cfda5a924a9ff497372a98c09a3acf6f54a9ff7fa7687ca78073e52a832eead"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-23T01:25:04.355409Z","signature_b64":"8GGFjydSeOKZoS99Q09P5oFFmrwbCLSpyRwigmQN1DK5HVpFEXLZHMcSSFmQTIpQMMnuD619iimHlOmMwbJ+Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"402c21ff6d50f61124a4f0f619038fc18f4f158289990eb722a071cce640a354","last_reissued_at":"2026-07-23T01:25:04.354496Z","signature_status":"signed_v1","first_computed_at":"2026-07-23T01:25:04.354496Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Abdullin Amir, Baotian Hu, Barkova Maria, Bingwei Zhang, Chang Liu, Chenwei Wu, Cuihua Hu, Dongfang Li, Dong Zhang, Fangxue Zhang, Haizhou Li, Hongzhang Yang, Igor Vasilyev, Jiajun Liu, Jian Meng, Jie Yu, Jing Li, Jinhua Zhou, Jinmin Xiang, Junjie Huang, Lin Hu, Linxin Yang, Linyuan Qiu, Li Zeng, Mingzhe Yang, Min Liu, Min Zhang, Nosov Ivan, Qian Chen, Ruoyu Sun, Shihao Hong, Shihao Zeng, Shijun Qin, Sihang Chen, Tao Guo, Taolue Chen, Tian Ding, Tianxiang Fang, Tolstykh Vasiliy, Ushakov Anton, Weihua Dai, Weijian Sun, Weiyan Kong, Wenhan Luo, Wenlong Jin, Xianfei Jin, Xiaodong Luo, Xin Chen, Xin Zhang, Xuhui Chen, Xutong Zhao, Yangbo Guo, Yingmeng Gao, Yiting Wang, Yucheng Xie, Zhengjun Yue, Zhengxuan Lu, Zheng Zhang, Zhichen Zhou, Zhihang Lin, Zhiquan Luo, Zhixiong Ning, Zhuohan Wang, Zirong Zeng, Ziwei Zhu","submitted_at":"2026-07-22T13:49:17Z","abstract_excerpt":"Full-parameter post-training of trillion-parameter-scale MoE models introduces substantial system-level challenges for large-scale distributed training, including severe memory pressure, non-overlapped communication overhead, and inefficient kernel execution. While most large-scale LLM training systems are built around GPU-based clusters, this report presents an end-to-end optimization practice on the Ascend NPU SuperPOD. Using the DeepSeek-V4 model family as the target workload, we develop a hierarchical optimization framework spanning model-level parallelism, computation-communication orches"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.20145","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2607.20145/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2607.20145","created_at":"2026-07-23T01:25:04.354978+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.20145v1","created_at":"2026-07-23T01:25:04.354978+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.20145","created_at":"2026-07-23T01:25:04.354978+00:00"},{"alias_kind":"pith_short_12","alias_value":"IAWCD73NKD3B","created_at":"2026-07-23T01:25:04.354978+00:00"},{"alias_kind":"pith_short_16","alias_value":"IAWCD73NKD3BCJFE","created_at":"2026-07-23T01:25:04.354978+00:00"},{"alias_kind":"pith_short_8","alias_value":"IAWCD73N","created_at":"2026-07-23T01:25:04.354978+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IAWCD73NKD3BCJFE6D3BSA4PYG","json":"https://pith.science/pith/IAWCD73NKD3BCJFE6D3BSA4PYG.json","graph_json":"https://pith.science/api/pith-number/IAWCD73NKD3BCJFE6D3BSA4PYG/graph.json","events_json":"https://pith.science/api/pith-number/IAWCD73NKD3BCJFE6D3BSA4PYG/events.json","paper":"https://pith.science/paper/IAWCD73N"},"agent_actions":{"view_html":"https://pith.science/pith/IAWCD73NKD3BCJFE6D3BSA4PYG","download_json":"https://pith.science/pith/IAWCD73NKD3BCJFE6D3BSA4PYG.json","view_paper":"https://pith.science/paper/IAWCD73N","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.20145&json=true","fetch_graph":"https://pith.science/api/pith-number/IAWCD73NKD3BCJFE6D3BSA4PYG/graph.json","fetch_events":"https://pith.science/api/pith-number/IAWCD73NKD3BCJFE6D3BSA4PYG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IAWCD73NKD3BCJFE6D3BSA4PYG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IAWCD73NKD3BCJFE6D3BSA4PYG/action/storage_attestation","attest_author":"https://pith.science/pith/IAWCD73NKD3BCJFE6D3BSA4PYG/action/author_attestation","sign_citation":"https://pith.science/pith/IAWCD73NKD3BCJFE6D3BSA4PYG/action/citation_signature","submit_replication":"https://pith.science/pith/IAWCD73NKD3BCJFE6D3BSA4PYG/action/replication_record"}},"created_at":"2026-07-23T01:25:04.354978+00:00","updated_at":"2026-07-23T01:25:04.354978+00:00"}