{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5SSMJBCKUYMOCBRJY6IOWNBAGL","short_pith_number":"pith:5SSMJBCK","schema_version":"1.0","canonical_sha256":"eca4c4844aa618e10629c790eb342032d8f4a9891f50f9cd4bc7fa8707c023b8","source":{"kind":"arxiv","id":"2606.12280","version":1},"attestation_state":"computed","paper":{"title":"Holding the FP8 Quality Ceiling at 8-Bit Weights and Activations: INT8 and GGUF Post-Training Quantization of Ideogram 4.0 for Consumer GPUs","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ali Asaria, Deep Gandhi, Tony Salomone","submitted_at":"2026-06-10T16:19:49Z","abstract_excerpt":"Post-training quantization lets large text-to-image diffusion transformers run on consumer GPUs, yet the hardware-specific trade-offs are seldom measured directly. We quantize Ideogram 4.0 - a 9.3B flow-matching diffusion transformer (DiT), shipped as two separate-weight copies of a single-stream 34-layer backbone for classifier-free guidance and conditioned by a Qwen3-VL-8B encoder - for Ampere RTX 3090 GPUs, which lack FP8 tensor cores. Our INT8 W8A8 recipe (per-channel weights, per-token dynamic activations, SmoothQuant, and mixed-precision protection of a small high-fragility layer set) ho"},"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":"2606.12280","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-10T16:19:49Z","cross_cats_sorted":[],"title_canon_sha256":"366bac06b37502131b961e1c8091e283c918d829a3c7d14fc1a6b2d3bd26fc88","abstract_canon_sha256":"3964dd3242024a9bb2e2d280a3e32212e5a1ea5b0981c62c59178f7ebb64c9e9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-11T01:10:59.392136Z","signature_b64":"Hmk7FrqVOkoTiZp+Gup5EFYZTo2kzYweEQLF+ZNr918MZfxZTnVqlkPlWxX97pFNsk3t1slJb7YC/IE8rj0LDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eca4c4844aa618e10629c790eb342032d8f4a9891f50f9cd4bc7fa8707c023b8","last_reissued_at":"2026-06-11T01:10:59.391392Z","signature_status":"signed_v1","first_computed_at":"2026-06-11T01:10:59.391392Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Holding the FP8 Quality Ceiling at 8-Bit Weights and Activations: INT8 and GGUF Post-Training Quantization of Ideogram 4.0 for Consumer GPUs","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ali Asaria, Deep Gandhi, Tony Salomone","submitted_at":"2026-06-10T16:19:49Z","abstract_excerpt":"Post-training quantization lets large text-to-image diffusion transformers run on consumer GPUs, yet the hardware-specific trade-offs are seldom measured directly. We quantize Ideogram 4.0 - a 9.3B flow-matching diffusion transformer (DiT), shipped as two separate-weight copies of a single-stream 34-layer backbone for classifier-free guidance and conditioned by a Qwen3-VL-8B encoder - for Ampere RTX 3090 GPUs, which lack FP8 tensor cores. Our INT8 W8A8 recipe (per-channel weights, per-token dynamic activations, SmoothQuant, and mixed-precision protection of a small high-fragility layer set) ho"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.12280","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/2606.12280/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":"2606.12280","created_at":"2026-06-11T01:10:59.391509+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.12280v1","created_at":"2026-06-11T01:10:59.391509+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.12280","created_at":"2026-06-11T01:10:59.391509+00:00"},{"alias_kind":"pith_short_12","alias_value":"5SSMJBCKUYMO","created_at":"2026-06-11T01:10:59.391509+00:00"},{"alias_kind":"pith_short_16","alias_value":"5SSMJBCKUYMOCBRJ","created_at":"2026-06-11T01:10:59.391509+00:00"},{"alias_kind":"pith_short_8","alias_value":"5SSMJBCK","created_at":"2026-06-11T01:10:59.391509+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/5SSMJBCKUYMOCBRJY6IOWNBAGL","json":"https://pith.science/pith/5SSMJBCKUYMOCBRJY6IOWNBAGL.json","graph_json":"https://pith.science/api/pith-number/5SSMJBCKUYMOCBRJY6IOWNBAGL/graph.json","events_json":"https://pith.science/api/pith-number/5SSMJBCKUYMOCBRJY6IOWNBAGL/events.json","paper":"https://pith.science/paper/5SSMJBCK"},"agent_actions":{"view_html":"https://pith.science/pith/5SSMJBCKUYMOCBRJY6IOWNBAGL","download_json":"https://pith.science/pith/5SSMJBCKUYMOCBRJY6IOWNBAGL.json","view_paper":"https://pith.science/paper/5SSMJBCK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.12280&json=true","fetch_graph":"https://pith.science/api/pith-number/5SSMJBCKUYMOCBRJY6IOWNBAGL/graph.json","fetch_events":"https://pith.science/api/pith-number/5SSMJBCKUYMOCBRJY6IOWNBAGL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5SSMJBCKUYMOCBRJY6IOWNBAGL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5SSMJBCKUYMOCBRJY6IOWNBAGL/action/storage_attestation","attest_author":"https://pith.science/pith/5SSMJBCKUYMOCBRJY6IOWNBAGL/action/author_attestation","sign_citation":"https://pith.science/pith/5SSMJBCKUYMOCBRJY6IOWNBAGL/action/citation_signature","submit_replication":"https://pith.science/pith/5SSMJBCKUYMOCBRJY6IOWNBAGL/action/replication_record"}},"created_at":"2026-06-11T01:10:59.391509+00:00","updated_at":"2026-06-11T01:10:59.391509+00:00"}