{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:TTFDTLQCDPEVGVZNOP67WAXJDM","short_pith_number":"pith:TTFDTLQC","schema_version":"1.0","canonical_sha256":"9cca39ae021bc953572d73fdfb02e91b1c8f410eacd7b799be4f617be216e1b3","source":{"kind":"arxiv","id":"2309.10818","version":3},"attestation_state":"computed","paper":{"title":"SlimPajama-DC: Understanding Data Combinations for LLM Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bowen Tan, Daria Soboleva, Eric Xing, Hongyi Wang, Joel Hestness, Liqun Ma, Natalia Vassilieva, Tianhua Tao, Willie Neiswanger, Zhengzhong Liu, Zhiqiang Shen","submitted_at":"2023-09-19T17:59:54Z","abstract_excerpt":"This paper aims to understand the impacts of various data combinations (e.g., web text, Wikipedia, GitHub, books) on the pretraining of large language models using SlimPajama. SlimPajama is a rigorously deduplicated, multi-source dataset, which has been refined and further deduplicated to 627B tokens from the extensive 1.2T token RedPajama dataset contributed by Together. We have termed our research as SlimPajama-DC, an empirical analysis designed to uncover fundamental characteristics and best practices associated with employing SlimPajama in the training of large language models. During our "},"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":"2309.10818","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-19T17:59:54Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2c33e708f0677ddf592c361a7fe7e0739750328a6f50131aad4d123ce3f64a2e","abstract_canon_sha256":"b8399c163c6d413a76b10fa64cd6290a147d2b0ee67221992de46b448c02defe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:17:10.676999Z","signature_b64":"HqKQkgXMP8ZrI355CzwUiOM9oRm0gTvbbf7HjSs97+N5rCmko11Dwj/NBsklLEz1/1QQ9t5aTwpzA5Up0ySCAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9cca39ae021bc953572d73fdfb02e91b1c8f410eacd7b799be4f617be216e1b3","last_reissued_at":"2026-07-05T08:17:10.676274Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:17:10.676274Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SlimPajama-DC: Understanding Data Combinations for LLM Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bowen Tan, Daria Soboleva, Eric Xing, Hongyi Wang, Joel Hestness, Liqun Ma, Natalia Vassilieva, Tianhua Tao, Willie Neiswanger, Zhengzhong Liu, Zhiqiang Shen","submitted_at":"2023-09-19T17:59:54Z","abstract_excerpt":"This paper aims to understand the impacts of various data combinations (e.g., web text, Wikipedia, GitHub, books) on the pretraining of large language models using SlimPajama. SlimPajama is a rigorously deduplicated, multi-source dataset, which has been refined and further deduplicated to 627B tokens from the extensive 1.2T token RedPajama dataset contributed by Together. We have termed our research as SlimPajama-DC, an empirical analysis designed to uncover fundamental characteristics and best practices associated with employing SlimPajama in the training of large language models. During our "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.10818","kind":"arxiv","version":3},"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/2309.10818/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":"2309.10818","created_at":"2026-07-05T08:17:10.676367+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.10818v3","created_at":"2026-07-05T08:17:10.676367+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.10818","created_at":"2026-07-05T08:17:10.676367+00:00"},{"alias_kind":"pith_short_12","alias_value":"TTFDTLQCDPEV","created_at":"2026-07-05T08:17:10.676367+00:00"},{"alias_kind":"pith_short_16","alias_value":"TTFDTLQCDPEVGVZN","created_at":"2026-07-05T08:17:10.676367+00:00"},{"alias_kind":"pith_short_8","alias_value":"TTFDTLQC","created_at":"2026-07-05T08:17:10.676367+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.02559","citing_title":"From Layers to Submodules: Rethinking Granularity in Replacement-Based LLM Compression","ref_index":113,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15913","citing_title":"Towards Generalization of Block Attention via Automatic Segmentation and Block Distillation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2411.10915","citing_title":"Bias in Large Language Models: Origin, Evaluation, and Mitigation","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2501.01046","citing_title":"SEDD: Scalable and Efficient Dataset Deduplication with GPUs","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2502.12120","citing_title":"LLMs on the Line: Data Determines Loss-to-Loss Scaling Laws","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2505.12942","citing_title":"A3 : an Analytical Low-Rank Approximation Framework for Attention","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15913","citing_title":"Towards Generalization of Block Attention via Automatic Segmentation and Block Distillation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15913","citing_title":"Towards Generalization of Block Attention via Automatic Segmentation and Block Distillation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2507.15640","citing_title":"Data Mixing Agent: Learning to Re-weight Domains for Continual Pre-training","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10119","citing_title":"Refresh-Scaling the Memory of Balanced Adam","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09011","citing_title":"A Geometric Perspective on Next-Token Prediction in Large Language Models: Three Emerging Phases","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10119","citing_title":"Refresh-Scaling the Memory of Balanced Adam","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2303.18223","citing_title":"A Survey of Large Language Models","ref_index":253,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TTFDTLQCDPEVGVZNOP67WAXJDM","json":"https://pith.science/pith/TTFDTLQCDPEVGVZNOP67WAXJDM.json","graph_json":"https://pith.science/api/pith-number/TTFDTLQCDPEVGVZNOP67WAXJDM/graph.json","events_json":"https://pith.science/api/pith-number/TTFDTLQCDPEVGVZNOP67WAXJDM/events.json","paper":"https://pith.science/paper/TTFDTLQC"},"agent_actions":{"view_html":"https://pith.science/pith/TTFDTLQCDPEVGVZNOP67WAXJDM","download_json":"https://pith.science/pith/TTFDTLQCDPEVGVZNOP67WAXJDM.json","view_paper":"https://pith.science/paper/TTFDTLQC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.10818&json=true","fetch_graph":"https://pith.science/api/pith-number/TTFDTLQCDPEVGVZNOP67WAXJDM/graph.json","fetch_events":"https://pith.science/api/pith-number/TTFDTLQCDPEVGVZNOP67WAXJDM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TTFDTLQCDPEVGVZNOP67WAXJDM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TTFDTLQCDPEVGVZNOP67WAXJDM/action/storage_attestation","attest_author":"https://pith.science/pith/TTFDTLQCDPEVGVZNOP67WAXJDM/action/author_attestation","sign_citation":"https://pith.science/pith/TTFDTLQCDPEVGVZNOP67WAXJDM/action/citation_signature","submit_replication":"https://pith.science/pith/TTFDTLQCDPEVGVZNOP67WAXJDM/action/replication_record"}},"created_at":"2026-07-05T08:17:10.676367+00:00","updated_at":"2026-07-05T08:17:10.676367+00:00"}