{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:DC5PCRAQNNJPU77QFI5SA5CMYS","short_pith_number":"pith:DC5PCRAQ","schema_version":"1.0","canonical_sha256":"18baf144106b52fa7ff02a3b20744cc4b3d57f0d3c023df93fb29df635fab324","source":{"kind":"arxiv","id":"2309.14327","version":3},"attestation_state":"computed","paper":{"title":"DeepSpeed-VisualChat: Multi-Round Multi-Image Interleave Chat via Multi-Modal Causal Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Ammar Ahmad Awan, Conglong Li, Heyang Qin, Minjia Zhang, Olatunji Ruwase, Samyam Rajbhandari, Xiaoxia Wu, Yuxiong He, Zhewei Yao","submitted_at":"2023-09-25T17:53:29Z","abstract_excerpt":"Most of the existing multi-modal models, hindered by their incapacity to adeptly manage interleaved image-and-text inputs in multi-image, multi-round dialogues, face substantial constraints in resource allocation for training and data accessibility, impacting their adaptability and scalability across varied interaction realms. To address this, we present the DeepSpeed-VisualChat framework, designed to optimize Large Language Models (LLMs) by incorporating multi-modal capabilities, with a focus on enhancing the proficiency of Large Vision and Language Models in handling interleaved inputs. 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":"2309.14327","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-09-25T17:53:29Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"aaf5b0ec3f9803b84b74bd1928e2bff6f441e2842b859fb2cc47bd7b9fad0ba5","abstract_canon_sha256":"a71924a8448da555befefc487181d1411bffdcc71427aab32024543fd5169bcf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:17:54.945323Z","signature_b64":"nGFHUaGRTJM9CitFWlyzBw2iC3kenxmY1MlkwVQGY67merbeJaTi4KWE4BQDjR3sWCjf601IhTG8x5hwRqdTCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"18baf144106b52fa7ff02a3b20744cc4b3d57f0d3c023df93fb29df635fab324","last_reissued_at":"2026-07-05T07:17:54.944834Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:17:54.944834Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DeepSpeed-VisualChat: Multi-Round Multi-Image Interleave Chat via Multi-Modal Causal Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Ammar Ahmad Awan, Conglong Li, Heyang Qin, Minjia Zhang, Olatunji Ruwase, Samyam Rajbhandari, Xiaoxia Wu, Yuxiong He, Zhewei Yao","submitted_at":"2023-09-25T17:53:29Z","abstract_excerpt":"Most of the existing multi-modal models, hindered by their incapacity to adeptly manage interleaved image-and-text inputs in multi-image, multi-round dialogues, face substantial constraints in resource allocation for training and data accessibility, impacting their adaptability and scalability across varied interaction realms. To address this, we present the DeepSpeed-VisualChat framework, designed to optimize Large Language Models (LLMs) by incorporating multi-modal capabilities, with a focus on enhancing the proficiency of Large Vision and Language Models in handling interleaved inputs. Our "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.14327","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/2309.14327/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":"2309.14327","created_at":"2026-07-05T07:17:54.944888+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.14327v3","created_at":"2026-07-05T07:17:54.944888+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.14327","created_at":"2026-07-05T07:17:54.944888+00:00"},{"alias_kind":"pith_short_12","alias_value":"DC5PCRAQNNJP","created_at":"2026-07-05T07:17:54.944888+00:00"},{"alias_kind":"pith_short_16","alias_value":"DC5PCRAQNNJPU77Q","created_at":"2026-07-05T07:17:54.944888+00:00"},{"alias_kind":"pith_short_8","alias_value":"DC5PCRAQ","created_at":"2026-07-05T07:17:54.944888+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2504.16030","citing_title":"LiveCC: Learning Video LLM with Streaming Speech Transcription at Scale","ref_index":93,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DC5PCRAQNNJPU77QFI5SA5CMYS","json":"https://pith.science/pith/DC5PCRAQNNJPU77QFI5SA5CMYS.json","graph_json":"https://pith.science/api/pith-number/DC5PCRAQNNJPU77QFI5SA5CMYS/graph.json","events_json":"https://pith.science/api/pith-number/DC5PCRAQNNJPU77QFI5SA5CMYS/events.json","paper":"https://pith.science/paper/DC5PCRAQ"},"agent_actions":{"view_html":"https://pith.science/pith/DC5PCRAQNNJPU77QFI5SA5CMYS","download_json":"https://pith.science/pith/DC5PCRAQNNJPU77QFI5SA5CMYS.json","view_paper":"https://pith.science/paper/DC5PCRAQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.14327&json=true","fetch_graph":"https://pith.science/api/pith-number/DC5PCRAQNNJPU77QFI5SA5CMYS/graph.json","fetch_events":"https://pith.science/api/pith-number/DC5PCRAQNNJPU77QFI5SA5CMYS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DC5PCRAQNNJPU77QFI5SA5CMYS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DC5PCRAQNNJPU77QFI5SA5CMYS/action/storage_attestation","attest_author":"https://pith.science/pith/DC5PCRAQNNJPU77QFI5SA5CMYS/action/author_attestation","sign_citation":"https://pith.science/pith/DC5PCRAQNNJPU77QFI5SA5CMYS/action/citation_signature","submit_replication":"https://pith.science/pith/DC5PCRAQNNJPU77QFI5SA5CMYS/action/replication_record"}},"created_at":"2026-07-05T07:17:54.944888+00:00","updated_at":"2026-07-05T07:17:54.944888+00:00"}