{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:LJSMJK2NKPP5LGLR2AEMCOZ6PT","short_pith_number":"pith:LJSMJK2N","schema_version":"1.0","canonical_sha256":"5a64c4ab4d53dfd59971d008c13b3e7cec121d68ce8c4dc9dc9727fc858de705","source":{"kind":"arxiv","id":"2010.03813","version":2},"attestation_state":"computed","paper":{"title":"On the importance of pre-training data volume for compact language models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Fran\\c{c}ois Fleuret, Martin d'Hoffschmidt, Vincent Micheli","submitted_at":"2020-10-08T07:40:21Z","abstract_excerpt":"Recent advances in language modeling have led to computationally intensive and resource-demanding state-of-the-art models. In an effort towards sustainable practices, we study the impact of pre-training data volume on compact language models. Multiple BERT-based models are trained on gradually increasing amounts of French text. Through fine-tuning on the French Question Answering Dataset (FQuAD), we observe that well-performing models are obtained with as little as 100 MB of text. In addition, we show that past critically low amounts of pre-training data, an intermediate pre-training step on t"},"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":"2010.03813","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-10-08T07:40:21Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ae68b01def93249a5f5e37ec63b1a7a56fd43bffd978784fb129ff077a5fa85f","abstract_canon_sha256":"96f53579918cdcf762e8d3d6b74dba49288c9f9e5e5a3e5f69385d71f7b779b9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:41:34.299173Z","signature_b64":"/R1LxIKuDIFnj+vIPtuhIDJVz286hZZ3RR2v7wF/vKdtNt7bHx/RZLoHrJRCNFo3yJuBwnX8TzAhN7aAN+jIDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5a64c4ab4d53dfd59971d008c13b3e7cec121d68ce8c4dc9dc9727fc858de705","last_reissued_at":"2026-07-05T01:41:34.298791Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:41:34.298791Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the importance of pre-training data volume for compact language models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Fran\\c{c}ois Fleuret, Martin d'Hoffschmidt, Vincent Micheli","submitted_at":"2020-10-08T07:40:21Z","abstract_excerpt":"Recent advances in language modeling have led to computationally intensive and resource-demanding state-of-the-art models. In an effort towards sustainable practices, we study the impact of pre-training data volume on compact language models. Multiple BERT-based models are trained on gradually increasing amounts of French text. Through fine-tuning on the French Question Answering Dataset (FQuAD), we observe that well-performing models are obtained with as little as 100 MB of text. In addition, we show that past critically low amounts of pre-training data, an intermediate pre-training step on t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.03813","kind":"arxiv","version":2},"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/2010.03813/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":"2010.03813","created_at":"2026-07-05T01:41:34.298847+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.03813v2","created_at":"2026-07-05T01:41:34.298847+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.03813","created_at":"2026-07-05T01:41:34.298847+00:00"},{"alias_kind":"pith_short_12","alias_value":"LJSMJK2NKPP5","created_at":"2026-07-05T01:41:34.298847+00:00"},{"alias_kind":"pith_short_16","alias_value":"LJSMJK2NKPP5LGLR","created_at":"2026-07-05T01:41:34.298847+00:00"},{"alias_kind":"pith_short_8","alias_value":"LJSMJK2N","created_at":"2026-07-05T01:41:34.298847+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/LJSMJK2NKPP5LGLR2AEMCOZ6PT","json":"https://pith.science/pith/LJSMJK2NKPP5LGLR2AEMCOZ6PT.json","graph_json":"https://pith.science/api/pith-number/LJSMJK2NKPP5LGLR2AEMCOZ6PT/graph.json","events_json":"https://pith.science/api/pith-number/LJSMJK2NKPP5LGLR2AEMCOZ6PT/events.json","paper":"https://pith.science/paper/LJSMJK2N"},"agent_actions":{"view_html":"https://pith.science/pith/LJSMJK2NKPP5LGLR2AEMCOZ6PT","download_json":"https://pith.science/pith/LJSMJK2NKPP5LGLR2AEMCOZ6PT.json","view_paper":"https://pith.science/paper/LJSMJK2N","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.03813&json=true","fetch_graph":"https://pith.science/api/pith-number/LJSMJK2NKPP5LGLR2AEMCOZ6PT/graph.json","fetch_events":"https://pith.science/api/pith-number/LJSMJK2NKPP5LGLR2AEMCOZ6PT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LJSMJK2NKPP5LGLR2AEMCOZ6PT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LJSMJK2NKPP5LGLR2AEMCOZ6PT/action/storage_attestation","attest_author":"https://pith.science/pith/LJSMJK2NKPP5LGLR2AEMCOZ6PT/action/author_attestation","sign_citation":"https://pith.science/pith/LJSMJK2NKPP5LGLR2AEMCOZ6PT/action/citation_signature","submit_replication":"https://pith.science/pith/LJSMJK2NKPP5LGLR2AEMCOZ6PT/action/replication_record"}},"created_at":"2026-07-05T01:41:34.298847+00:00","updated_at":"2026-07-05T01:41:34.298847+00:00"}