{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:3YFCD7HFDQ65L3BSEBA2WZUEKE","short_pith_number":"pith:3YFCD7HF","canonical_record":{"source":{"id":"2305.18627","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2023-05-29T21:32:15Z","cross_cats_sorted":["cs.DC","stat.ML"],"title_canon_sha256":"a7302377cc1107ca7164376be23b301b6a7114fddc29a2402d1fe86baeb3cc08","abstract_canon_sha256":"5300510a8b2aeea5ac4ea6afb314ea43e207f2345e467c295c8f320fe41d74a7"},"schema_version":"1.0"},"canonical_sha256":"de0a21fce51c3dd5ec322041ab6684511a35368699ca13162c997d657e082bf3","source":{"kind":"arxiv","id":"2305.18627","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.18627","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"arxiv_version","alias_value":"2305.18627v2","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.18627","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"pith_short_12","alias_value":"3YFCD7HFDQ65","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"pith_short_16","alias_value":"3YFCD7HFDQ65L3BS","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"pith_short_8","alias_value":"3YFCD7HF","created_at":"2026-07-05T11:44:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:3YFCD7HFDQ65L3BSEBA2WZUEKE","target":"record","payload":{"canonical_record":{"source":{"id":"2305.18627","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2023-05-29T21:32:15Z","cross_cats_sorted":["cs.DC","stat.ML"],"title_canon_sha256":"a7302377cc1107ca7164376be23b301b6a7114fddc29a2402d1fe86baeb3cc08","abstract_canon_sha256":"5300510a8b2aeea5ac4ea6afb314ea43e207f2345e467c295c8f320fe41d74a7"},"schema_version":"1.0"},"canonical_sha256":"de0a21fce51c3dd5ec322041ab6684511a35368699ca13162c997d657e082bf3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:44:37.886864Z","signature_b64":"Slp+wxKfT7sTCquFmv+h/bI4DN+FcqlfY0XyvGJJ7RGWTvdviOlAUpkDLuolyXkBzEEZ8TsmUHFpH8r/WwtCBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"de0a21fce51c3dd5ec322041ab6684511a35368699ca13162c997d657e082bf3","last_reissued_at":"2026-07-05T11:44:37.886352Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:44:37.886352Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2305.18627","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:44:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AFGc3M8MpA3I8g/iWNs26N9rsUO82HnRD79ApM7Uw0TR9gpo2jxsyDNbwrWoyaS8HR9aLKycC0vbHV79353/Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T18:28:15.663172Z"},"content_sha256":"c8e647a97e27157f635c22b38d3312543e1b98c19307341b6421ecb8f662fea7","schema_version":"1.0","event_id":"sha256:c8e647a97e27157f635c22b38d3312543e1b98c19307341b6421ecb8f662fea7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:3YFCD7HFDQ65L3BSEBA2WZUEKE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Quantize Once, Train Fast: Allreduce-Compatible Compression with Provable Guarantees","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.DC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jihao Xin, Marco Canini, Peter Richt\\'arik, Samuel Horv\\'ath","submitted_at":"2023-05-29T21:32:15Z","abstract_excerpt":"Distributed training enables large-scale deep learning, but suffers from high communication overhead, especially as models and datasets grow. Gradient compression, particularly quantization, is a promising approach to mitigate this bottleneck. However, existing quantization schemes are often incompatible with Allreduce, the dominant communication primitive in distributed deep learning, and many prior solutions rely on heuristics without theoretical guarantees. We introduce Global-QSGD, an Allreduce-compatible gradient quantization method that leverages global norm scaling to reduce communicati"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.18627","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/2305.18627/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:44:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/JRElo5wmAZtD//1dySpFgbE/Cv7kPyzigBFDUngnUXcmFqgdnQ8n8SjnpZsIj+cihLJrZ+kjs3Chnu+r7N1Aw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T18:28:15.663707Z"},"content_sha256":"cc8cf72e345bfce4f38ddafabb2d642dcebebcaa1bde88997295550d4d3f0008","schema_version":"1.0","event_id":"sha256:cc8cf72e345bfce4f38ddafabb2d642dcebebcaa1bde88997295550d4d3f0008"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3YFCD7HFDQ65L3BSEBA2WZUEKE/bundle.json","state_url":"https://pith.science/pith/3YFCD7HFDQ65L3BSEBA2WZUEKE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3YFCD7HFDQ65L3BSEBA2WZUEKE/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-04T18:28:15Z","links":{"resolver":"https://pith.science/pith/3YFCD7HFDQ65L3BSEBA2WZUEKE","bundle":"https://pith.science/pith/3YFCD7HFDQ65L3BSEBA2WZUEKE/bundle.json","state":"https://pith.science/pith/3YFCD7HFDQ65L3BSEBA2WZUEKE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3YFCD7HFDQ65L3BSEBA2WZUEKE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:3YFCD7HFDQ65L3BSEBA2WZUEKE","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"5300510a8b2aeea5ac4ea6afb314ea43e207f2345e467c295c8f320fe41d74a7","cross_cats_sorted":["cs.DC","stat.ML"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2023-05-29T21:32:15Z","title_canon_sha256":"a7302377cc1107ca7164376be23b301b6a7114fddc29a2402d1fe86baeb3cc08"},"schema_version":"1.0","source":{"id":"2305.18627","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.18627","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"arxiv_version","alias_value":"2305.18627v2","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.18627","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"pith_short_12","alias_value":"3YFCD7HFDQ65","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"pith_short_16","alias_value":"3YFCD7HFDQ65L3BS","created_at":"2026-07-05T11:44:37Z"},{"alias_kind":"pith_short_8","alias_value":"3YFCD7HF","created_at":"2026-07-05T11:44:37Z"}],"graph_snapshots":[{"event_id":"sha256:cc8cf72e345bfce4f38ddafabb2d642dcebebcaa1bde88997295550d4d3f0008","target":"graph","created_at":"2026-07-05T11:44:37Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2305.18627/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Distributed training enables large-scale deep learning, but suffers from high communication overhead, especially as models and datasets grow. Gradient compression, particularly quantization, is a promising approach to mitigate this bottleneck. However, existing quantization schemes are often incompatible with Allreduce, the dominant communication primitive in distributed deep learning, and many prior solutions rely on heuristics without theoretical guarantees. We introduce Global-QSGD, an Allreduce-compatible gradient quantization method that leverages global norm scaling to reduce communicati","authors_text":"Jihao Xin, Marco Canini, Peter Richt\\'arik, Samuel Horv\\'ath","cross_cats":["cs.DC","stat.ML"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2023-05-29T21:32:15Z","title":"Quantize Once, Train Fast: Allreduce-Compatible Compression with Provable Guarantees"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.18627","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:c8e647a97e27157f635c22b38d3312543e1b98c19307341b6421ecb8f662fea7","target":"record","created_at":"2026-07-05T11:44:37Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"5300510a8b2aeea5ac4ea6afb314ea43e207f2345e467c295c8f320fe41d74a7","cross_cats_sorted":["cs.DC","stat.ML"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2023-05-29T21:32:15Z","title_canon_sha256":"a7302377cc1107ca7164376be23b301b6a7114fddc29a2402d1fe86baeb3cc08"},"schema_version":"1.0","source":{"id":"2305.18627","kind":"arxiv","version":2}},"canonical_sha256":"de0a21fce51c3dd5ec322041ab6684511a35368699ca13162c997d657e082bf3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"de0a21fce51c3dd5ec322041ab6684511a35368699ca13162c997d657e082bf3","first_computed_at":"2026-07-05T11:44:37.886352Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:44:37.886352Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Slp+wxKfT7sTCquFmv+h/bI4DN+FcqlfY0XyvGJJ7RGWTvdviOlAUpkDLuolyXkBzEEZ8TsmUHFpH8r/WwtCBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:44:37.886864Z","signed_message":"canonical_sha256_bytes"},"source_id":"2305.18627","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c8e647a97e27157f635c22b38d3312543e1b98c19307341b6421ecb8f662fea7","sha256:cc8cf72e345bfce4f38ddafabb2d642dcebebcaa1bde88997295550d4d3f0008"],"state_sha256":"e0400a8e16224f4b57d0a78d7cd116337d5753cabd98ee119604599f772dcd28"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8gTCYQZJCCbnlN5RX8AbHPMxNOIeK2+12TbTWLkNbQ1jqX7BCzSDVtUC8xOp6GV1i1eHXwubM9OVn7PLoHPdDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T18:28:15.667447Z","bundle_sha256":"0b333a7b9be54b9fba21d7f683e50a0154a264232e58f5c4f28a6daa4bfcfb5e"}}