{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:MEI7DRMFATRBFJWULL2XB6HLRR","short_pith_number":"pith:MEI7DRMF","schema_version":"1.0","canonical_sha256":"6111f1c58504e212a6d45af570f8eb8c43440ef6c70f4aa2227a57358654aba9","source":{"kind":"arxiv","id":"2210.11803","version":1},"attestation_state":"computed","paper":{"title":"Revisiting Checkpoint Averaging for Neural Machine Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Christian Herold, Hermann Ney, Yingbo Gao, Zijian Yang","submitted_at":"2022-10-21T08:29:23Z","abstract_excerpt":"Checkpoint averaging is a simple and effective method to boost the performance of converged neural machine translation models. The calculation is cheap to perform and the fact that the translation improvement almost comes for free, makes it widely adopted in neural machine translation research. Despite the popularity, the method itself simply takes the mean of the model parameters from several checkpoints, the selection of which is mostly based on empirical recipes without many justifications. In this work, we revisit the concept of checkpoint averaging and consider several extensions. Specifi"},"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":"2210.11803","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-10-21T08:29:23Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"76f84763e002f643aee82d95d6b58feb6c2873f99ed34c038931903fd18ff0f5","abstract_canon_sha256":"f32758ca0d5d4da93d160e2b2c2123ae941418aacaa1f34f0269d6ef042ea84f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:08:57.828550Z","signature_b64":"K/8QsmDCxu8riiBoiqND/W3K6DK7oDlKVn0B3O+Swr/mZGW2qp4Whe7Mj/OP3/45QgBBWsuIpVqGxWkENkgJDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6111f1c58504e212a6d45af570f8eb8c43440ef6c70f4aa2227a57358654aba9","last_reissued_at":"2026-07-05T05:08:57.828129Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:08:57.828129Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Revisiting Checkpoint Averaging for Neural Machine Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Christian Herold, Hermann Ney, Yingbo Gao, Zijian Yang","submitted_at":"2022-10-21T08:29:23Z","abstract_excerpt":"Checkpoint averaging is a simple and effective method to boost the performance of converged neural machine translation models. The calculation is cheap to perform and the fact that the translation improvement almost comes for free, makes it widely adopted in neural machine translation research. Despite the popularity, the method itself simply takes the mean of the model parameters from several checkpoints, the selection of which is mostly based on empirical recipes without many justifications. In this work, we revisit the concept of checkpoint averaging and consider several extensions. Specifi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.11803","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/2210.11803/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":"2210.11803","created_at":"2026-07-05T05:08:57.828185+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.11803v1","created_at":"2026-07-05T05:08:57.828185+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.11803","created_at":"2026-07-05T05:08:57.828185+00:00"},{"alias_kind":"pith_short_12","alias_value":"MEI7DRMFATRB","created_at":"2026-07-05T05:08:57.828185+00:00"},{"alias_kind":"pith_short_16","alias_value":"MEI7DRMFATRBFJWU","created_at":"2026-07-05T05:08:57.828185+00:00"},{"alias_kind":"pith_short_8","alias_value":"MEI7DRMF","created_at":"2026-07-05T05:08:57.828185+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22917","citing_title":"GRAIN: Group Aggregation via Min-Norm Objective","ref_index":70,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MEI7DRMFATRBFJWULL2XB6HLRR","json":"https://pith.science/pith/MEI7DRMFATRBFJWULL2XB6HLRR.json","graph_json":"https://pith.science/api/pith-number/MEI7DRMFATRBFJWULL2XB6HLRR/graph.json","events_json":"https://pith.science/api/pith-number/MEI7DRMFATRBFJWULL2XB6HLRR/events.json","paper":"https://pith.science/paper/MEI7DRMF"},"agent_actions":{"view_html":"https://pith.science/pith/MEI7DRMFATRBFJWULL2XB6HLRR","download_json":"https://pith.science/pith/MEI7DRMFATRBFJWULL2XB6HLRR.json","view_paper":"https://pith.science/paper/MEI7DRMF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.11803&json=true","fetch_graph":"https://pith.science/api/pith-number/MEI7DRMFATRBFJWULL2XB6HLRR/graph.json","fetch_events":"https://pith.science/api/pith-number/MEI7DRMFATRBFJWULL2XB6HLRR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MEI7DRMFATRBFJWULL2XB6HLRR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MEI7DRMFATRBFJWULL2XB6HLRR/action/storage_attestation","attest_author":"https://pith.science/pith/MEI7DRMFATRBFJWULL2XB6HLRR/action/author_attestation","sign_citation":"https://pith.science/pith/MEI7DRMFATRBFJWULL2XB6HLRR/action/citation_signature","submit_replication":"https://pith.science/pith/MEI7DRMFATRBFJWULL2XB6HLRR/action/replication_record"}},"created_at":"2026-07-05T05:08:57.828185+00:00","updated_at":"2026-07-05T05:08:57.828185+00:00"}