{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:T5KOPKEZIJ6S2LTOYXOLAM5LEF","short_pith_number":"pith:T5KOPKEZ","schema_version":"1.0","canonical_sha256":"9f54e7a899427d2d2e6ec5dcb033ab216be64712864a2d2f473692905989a9c7","source":{"kind":"arxiv","id":"1909.06695","version":1},"attestation_state":"computed","paper":{"title":"Ouroboros: On Accelerating Training of Transformer-Based Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Heng Huang, Lawrence Carin, Qian Yang, Wenlin Wang, Zhouyuan Huo","submitted_at":"2019-09-14T23:21:56Z","abstract_excerpt":"Language models are essential for natural language processing (NLP) tasks, such as machine translation and text summarization. Remarkable performance has been demonstrated recently across many NLP domains via a Transformer-based language model with over a billion parameters, verifying the benefits of model size. Model parallelism is required if a model is too large to fit in a single computing device. Current methods for model parallelism either suffer from backward locking in backpropagation or are not applicable to language models. We propose the first model-parallel algorithm that speeds th"},"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":"1909.06695","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-09-14T23:21:56Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"4d82a6dac14dec501e2f810d5bd037700d9c90d9e412024e1d2d1a4c5d3d7e1b","abstract_canon_sha256":"f67a7da9ec1557d4bcdb92241a3a2dbf8f302d96f7e6214d9ad4deba54a472c9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:04:48.744024Z","signature_b64":"hNVzwAHGoXXq1thBR0t+sVWLNDWR9hZnfbsxIFC2h+fNBN8fXXh1eGEy2eLiXq2EmhNCQvwZXkDREHUNXE12Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9f54e7a899427d2d2e6ec5dcb033ab216be64712864a2d2f473692905989a9c7","last_reissued_at":"2026-07-05T00:04:48.743399Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:04:48.743399Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Ouroboros: On Accelerating Training of Transformer-Based Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Heng Huang, Lawrence Carin, Qian Yang, Wenlin Wang, Zhouyuan Huo","submitted_at":"2019-09-14T23:21:56Z","abstract_excerpt":"Language models are essential for natural language processing (NLP) tasks, such as machine translation and text summarization. Remarkable performance has been demonstrated recently across many NLP domains via a Transformer-based language model with over a billion parameters, verifying the benefits of model size. Model parallelism is required if a model is too large to fit in a single computing device. Current methods for model parallelism either suffer from backward locking in backpropagation or are not applicable to language models. We propose the first model-parallel algorithm that speeds th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1909.06695","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/1909.06695/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":"1909.06695","created_at":"2026-07-05T00:04:48.743465+00:00"},{"alias_kind":"arxiv_version","alias_value":"1909.06695v1","created_at":"2026-07-05T00:04:48.743465+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1909.06695","created_at":"2026-07-05T00:04:48.743465+00:00"},{"alias_kind":"pith_short_12","alias_value":"T5KOPKEZIJ6S","created_at":"2026-07-05T00:04:48.743465+00:00"},{"alias_kind":"pith_short_16","alias_value":"T5KOPKEZIJ6S2LTO","created_at":"2026-07-05T00:04:48.743465+00:00"},{"alias_kind":"pith_short_8","alias_value":"T5KOPKEZ","created_at":"2026-07-05T00:04:48.743465+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"1909.02625","citing_title":"On the Acceleration of Deep Learning Model Parallelism with Staleness","ref_index":38,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/T5KOPKEZIJ6S2LTOYXOLAM5LEF","json":"https://pith.science/pith/T5KOPKEZIJ6S2LTOYXOLAM5LEF.json","graph_json":"https://pith.science/api/pith-number/T5KOPKEZIJ6S2LTOYXOLAM5LEF/graph.json","events_json":"https://pith.science/api/pith-number/T5KOPKEZIJ6S2LTOYXOLAM5LEF/events.json","paper":"https://pith.science/paper/T5KOPKEZ"},"agent_actions":{"view_html":"https://pith.science/pith/T5KOPKEZIJ6S2LTOYXOLAM5LEF","download_json":"https://pith.science/pith/T5KOPKEZIJ6S2LTOYXOLAM5LEF.json","view_paper":"https://pith.science/paper/T5KOPKEZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1909.06695&json=true","fetch_graph":"https://pith.science/api/pith-number/T5KOPKEZIJ6S2LTOYXOLAM5LEF/graph.json","fetch_events":"https://pith.science/api/pith-number/T5KOPKEZIJ6S2LTOYXOLAM5LEF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/T5KOPKEZIJ6S2LTOYXOLAM5LEF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/T5KOPKEZIJ6S2LTOYXOLAM5LEF/action/storage_attestation","attest_author":"https://pith.science/pith/T5KOPKEZIJ6S2LTOYXOLAM5LEF/action/author_attestation","sign_citation":"https://pith.science/pith/T5KOPKEZIJ6S2LTOYXOLAM5LEF/action/citation_signature","submit_replication":"https://pith.science/pith/T5KOPKEZIJ6S2LTOYXOLAM5LEF/action/replication_record"}},"created_at":"2026-07-05T00:04:48.743465+00:00","updated_at":"2026-07-05T00:04:48.743465+00:00"}