From bf8f76505fdd25ea99770f262eb9371509bf226d Mon Sep 17 00:00:00 2001 From: Chris Dumas Date: Tue, 20 Oct 2026 13:43:38 +0000 Subject: [PATCH] Squashed 'vendor/dumas_h3_continuum/' content from commit d07ed31 git-subtree-dir: vendor/dumas_h3_continuum git-subtree-split: d07ed31b442d283d6a1740f875a843b9feda4c91 --- .comfyignore | 99 + .gitattributes | 3 + .github/workflows/publish_registry.yml | 79 + .github/workflows/publish_release.yml | 154 + .github/workflows/tests.yml | 42 + .gitignore | 6 + AGENTS.md | 42 + CHANGELOG.md | 374 ++ H3-Continuum-LLM-System-Prompt-v1.zip | Bin 0 -> 11595 bytes H3-Continuum-Skill-v1.zip | Bin 0 -> 17773 bytes INSTALL_JA.txt | 64 + LICENSE | 21 + MANIFEST.sha256 | 477 +++ PACKAGE_INFO_JA.txt | 231 + PACKAGE_VALIDATION.txt | 354 ++ PROJECT_STATE.md | 147 + README.md | 306 ++ README_JA.md | 279 ++ REGISTRY_MANIFEST.sha256 | 171 + THIRD_PARTY_NOTICES.md | 12 + WORKLOG.md | 189 + __init__.py | 32 + audio_compat.py | 29 + branch_provenance.py | 222 + compatibility.py | 291 ++ conditioning.py | 105 + constants.py | 123 + continuation.py | 169 + decode_cache/__init__.py | 2 + decode_cache/identity.py | 212 + decode_cache/nodes.py | 118 + decode_cache/service.py | 115 + decode_cache/store.py | 427 ++ docs/AB_TEST_LOG_JA.md | 291 ++ docs/ARCHITECTURE.md | 177 + docs/FAILURE_SEMANTICS_BASELINE.md | 93 + docs/FAILURE_SEMANTICS_DELTA_R5.md | 90 + docs/FUTURE_ROADMAP.md | 16 + docs/ISSUE31_SAMPLING_SIGNATURES.md | 66 + docs/KJ_PREVIEW_OVERRIDE_V398.md | 95 + docs/LOADER_PERSISTENCE_REPAIR.md | 28 + docs/N2B_BRANCH_PROVENANCE_CONTRACT_V1.md | 69 + docs/R0_RUNTIME_CHARACTERIZATION.md | 215 + docs/R6_NODE_FACADE_DELTA.md | 41 + docs/README_HISTORY.md | 905 ++++ docs/README_HISTORY_JA.md | 517 +++ docs/REGENERATE_INTENT_V396.md | 37 + docs/RELEASE_NOTES_V3910.md | 67 + docs/RELEASE_NOTES_V395.md | 21 + docs/RELEASE_NOTES_V396.md | 21 + docs/RELEASE_NOTES_V397.md | 30 + docs/RELEASE_NOTES_V398.md | 47 + docs/RELEASE_NOTES_V399.md | 69 + docs/REVIEW_RESTORATION_REPAIR.md | 27 + docs/SEAM_TEST_SOURCE_JA.md | 56 + docs/TEMPORARY_STORAGE_JA.md | 210 + docs/TIMELINE_VIDEO_EXPERIMENTAL.md | 83 + docs/TURBO_SAMPLE.md | 44 + docs/USER_GUIDE.md | 705 ++++ docs/USER_GUIDE_JA.md | 309 ++ docs/V2_1_4_UI_SIMPLIFICATION.md | 35 + docs/V2_1_5_UI_SERIALIZATION_STABILITY.md | 20 + docs/V2_1_6_UNIFIED_UI.md | 37 + docs/V34_DRIVING_AUDIO_VALIDATION.md | 62 + docs/V35_HIRES_FIX.md | 154 + ...INNER_MANUAL_PLAN_AND_CAPTURE_SCRIPT_JA.md | 251 ++ docs/V38_OPEN_INTEGRATION_CONTRACT.md | 84 + docs/V38_RELEASE_AND_MIGRATION.md | 44 + docs/V39_REFERENCE_INLINE.md | 20 + docs/VALIDATION_JA.md | 246 ++ docs/VALIDATION_RESULTS.md | 99 + docs/images/comfyui-manager-h3-continuum.png | Bin 0 -> 15786 bytes docs/images/h3-continuum-advanced.png | Bin 0 -> 57536 bytes docs/images/h3-continuum-detailed-report.png | Bin 0 -> 240877 bytes docs/images/h3-continuum-image-scale.png | Bin 0 -> 29142 bytes docs/images/h3-continuum-sampler.png | Bin 0 -> 72111 bytes docs/images/h3-continuum-v33-settings.png | Bin 0 -> 35120 bytes .../h3-continuum-v33-streamlined-ui.png | Bin 0 -> 260441 bytes docs/images/v31b-auto-resume.png | Bin 0 -> 23828 bytes docs/images/v34-driving-audio-connection.png | Bin 0 -> 130087 bytes docs/images/v34-regenerate-from.png | Bin 0 -> 62052 bytes docs/images/v34-sampler-spectrum-assemble.png | Bin 0 -> 443545 bytes .../v34-video-reference-separate-audio.png | Bin 0 -> 171435 bytes .../images/v34-video-reference-with-audio.png | Bin 0 -> 290069 bytes docs/images/v34-workflow-overview.png | Bin 0 -> 582557 bytes docs/images/v35-assemble-seam-node.png | Bin 0 -> 26229 bytes docs/images/v35-hires-fix-node.png | Bin 0 -> 1327813 bytes docs/images/v35-second-pass-node.png | Bin 0 -> 1074278 bytes .../v351-lbh-conditioning-bridge-flow.svg | 142 + .../images/v351-reference-audio-permanent.png | Bin 0 -> 119746 bytes docs/images/v351-video-guide-frames.png | Bin 0 -> 118799 bytes .../v352-stabilization-optimization.png | Bin 0 -> 1662426 bytes docs/images/v36-masked-av-continuation.png | Bin 0 -> 1494049 bytes docs/images/v36-sampler-node.png | Bin 0 -> 73907 bytes docs/images/v37-sampler-node.png | Bin 0 -> 1327172 bytes docs/images/v37-tail6-second-pass.png | Bin 0 -> 1567151 bytes docs/images/v38-manual/continue-selected.png | Bin 0 -> 7986 bytes .../v38-manual/first-last-image-bypass.png | Bin 0 -> 735453 bytes .../images/v38-manual/full-video-progress.png | Bin 0 -> 20071 bytes .../v38-manual/h3-continuum-v38-civitai.png | Bin 0 -> 1981326 bytes .../v38-manual/h3-continuum-v38-whats-new.png | Bin 0 -> 1649397 bytes .../v38-manual/input-video-audio-bypass.png | Bin 0 -> 289220 bytes .../live-continue-selected-steps4.png | Bin 0 -> 48851 bytes .../images/v38-manual/live-execute-button.png | Bin 0 -> 8050 bytes docs/images/v38-manual/live-first-review.png | Bin 0 -> 47104 bytes .../live-regenerate-from-chunk2.png | Bin 0 -> 26130 bytes .../v38-manual/live-render-history-steps4.png | Bin 0 -> 63274 bytes .../images/v38-manual/live-retry-selected.png | Bin 0 -> 33710 bytes .../v38-manual/live-review-actions-steps4.png | Bin 0 -> 48486 bytes .../v38-manual/live-review-settings.png | Bin 0 -> 109618 bytes .../live-sequence-complete-steps4.png | Bin 0 -> 39440 bytes .../live-settings-complete-steps4.png | Bin 0 -> 115023 bytes .../reference-images-three-bypass.png | Bin 0 -> 175642 bytes docs/images/v38-manual/return-to-review.png | Bin 0 -> 28843 bytes docs/images/v38-manual/review-actions.png | Bin 0 -> 25160 bytes docs/images/v38-manual/review-settings.png | Bin 0 -> 56481 bytes .../sampler-size-first-image-draft.png | Bin 0 -> 33235 bytes docs/images/v38-manual/seed-fixed.png | Bin 0 -> 18413 bytes .../size-source-first-image-tooltip.png | Bin 0 -> 17843 bytes .../v38-manual/size-source-manual-640.png | Bin 0 -> 20635 bytes docs/images/v38-manual/turbo-lora-choices.png | Bin 0 -> 130508 bytes .../v39/v38x2-feature-summary-historical.png | Bin 0 -> 590455 bytes docs/images/v39/v39-feature-summary.png | Bin 0 -> 1420878 bytes docs/images/v39/v39-reference-images-node.png | Bin 0 -> 116859 bytes docs/images/v39/v39-sampler-node.png | Bin 0 -> 114354 bytes driving_audio.py | 192 + examples/H3_CONTINUUM_V1_WIRING.md | 33 + examples/H3_CONTINUUM_V2_WIRING_JA.md | 68 + examples/H3_Continuum_V2_README.txt | 31 + examples/H3_Continuum_V2_reference.webp | Bin 0 -> 48238 bytes examples/MiniMax_H3_Continuum_V38X2.json | 2319 ++++++++++ examples/MiniMax_H3_Continuum_V39.json | 2315 ++++++++++ ...MiniMax_H3_Continuum_V39_Turbo_Sample.json | 2414 +++++++++++ .../settings/recommended_settings_v1.json | 24 + .../settings/recommended_settings_v2.json | 29 + examples/v2_prompt_list_example.txt | 5 + examples/v2_prompt_timeline_example.txt | 8 + ...3_Continuum_V38X2+Decode_Cache_Helper.json | 2319 ++++++++++ ...H3_Continuum_V38X2+Decode_Cache_Helper.zip | Bin 0 -> 7423 bytes .../workflows/MiniMax_H3_Continuum_V38X2.json | 2319 ++++++++++ .../workflows/MiniMax_H3_Continuum_V38X2.zip | Bin 0 -> 13422 bytes ...um_V38X2_Timeline_Experimental_Follow.json | 2321 ++++++++++ ...uum_V38X2_Timeline_Experimental_Follow.zip | Bin 0 -> 7461 bytes ...um_V38X2_Timeline_Experimental_Repeat.json | 2321 ++++++++++ ...uum_V38X2_Timeline_Experimental_Repeat.zip | Bin 0 -> 7462 bytes .../workflows/MiniMax_H3_Continuum_V38x.json | 2078 +++++++++ .../workflows/MiniMax_H3_Continuum_V38x.zip | Bin 0 -> 6790 bytes .../workflows/MiniMax_H3_Continuum_V39.json | 2315 ++++++++++ .../workflows/MiniMax_H3_Continuum_V39.zip | Bin 0 -> 6227 bytes ...Max_H3_Continuum_V39_Reference_Inline.json | 2342 ++++++++++ ...iMax_H3_Continuum_V39_Reference_Inline.zip | Bin 0 -> 7297 bytes ...MiniMax_H3_Continuum_V39_Turbo_Sample.json | 2414 +++++++++++ .../MiniMax_H3_Continuum_V39_Turbo_Sample.zip | Bin 0 -> 7720 bytes graph_contract.py | 210 + guide_timeline.py | 357 ++ hardening.py | 708 ++++ install_windows.bat | 13 + install_windows.ps1 | 96 + layout_adapter.py | 651 +++ media.py | 91 + metadata.ini | 4 + minimax-h3-continuum-dance-director.zip | Bin 0 -> 20103 bytes model_patch.py | 371 ++ nodes.py | 481 +++ preview_override_provider_public_v1.json | 19 + preview_override_provider_v1.json | 19 + preview_override_signature.py | 294 ++ public_api.py | 55 + pyproject.toml | 18 + pytest.ini | 3 + reference.py | 484 +++ reference_audio.py | 506 +++ reference_video.py | 381 ++ requirements.txt | 1 + run_storage.py | 3751 +++++++++++++++++ spectrum_provider_v029.json | 1 + spectrum_provider_v029_py312.json | 1 + spectrum_provider_v1.json | 1 + spectrum_provider_v1_py312.json | 1 + spectrum_signature.py | 242 ++ state.py | 125 + state_io.py | 161 + storage_api.py | 97 + storage_fs.py | 129 + storage_memory.py | 52 + storage_settings.py | 460 ++ temporal.py | 140 + tests/conftest.py | 55 + tests/decode_cache/conftest.py | 74 + .../helpers/cache_process_worker.py | 47 + tests/decode_cache/test_adapter_nodes.py | 273 ++ tests/decode_cache/test_identity.py | 164 + tests/decode_cache/test_process_isolation.py | 50 + tests/decode_cache/test_store.py | 295 ++ .../issue31_local_spectrum_signature.py.txt | 200 + .../issue31_phase1_model_options.py.txt | 19 + ...8_a3c_selective_refine_saved_workflow.json | 26 + tests/fixtures/v39_sampling_v5/project.json | 99 + ...61481ead25f91627af0-chunk-0002.safetensors | Bin 0 -> 124928 bytes ...342b114bd31b175225c-chunk-0001.safetensors | Bin 0 -> 110080 bytes .../revisions/d8352287737286e5/manifest.json | 726 ++++ tests/frontend_decode_cache_helper.mjs | 91 + tests/frontend_external_int_size.cjs | 130 + tests/frontend_issue30_ui.cjs | 131 + tests/frontend_reference_images_v39.cjs | 307 ++ tests/frontend_review_queue.cjs | 821 ++++ tests/frontend_storage_settings.mjs | 128 + tests/helpers/r5_process_crash_worker.py | 165 + tests/terminal_merge_diagnostics.py | 378 ++ tests/test_audit_boundary_repairs.py | 227 + tests/test_compatibility.py | 83 + tests/test_conditioning_modes.py | 37 + tests/test_context_diagnostics.py | 46 + tests/test_easy_load_audio.py | 216 + tests/test_easy_load_image.py | 269 ++ tests/test_easy_load_video.py | 345 ++ tests/test_external_int_size.py | 58 + tests/test_gpu_diagnostic_runner.py | 81 + tests/test_issue13_r25_tools.py | 313 ++ tests/test_issue13_r28_tools.py | 213 + tests/test_issue13_r2_tools.py | 139 + tests/test_issue30_frontend.py | 54 + tests/test_issue31_attention_signature.py | 381 ++ ...test_issue31_preview_override_signature.py | 328 ++ tests/test_issue31_spectrum_signature.py | 329 ++ tests/test_layout_adapter.py | 188 + tests/test_loader_store_roundtrip.py | 116 + tests/test_m3_conditioning_display.py | 59 + tests/test_model_patch.py | 327 ++ tests/test_nodes_logic.py | 95 + tests/test_p20_file_backed_tensor_poc.py | 515 +++ tests/test_p21_memory_attribution.py | 585 +++ tests/test_p22_file_backed_buffer.py | 269 ++ tests/test_p23_assembly_v35.py | 796 ++++ tests/test_p23_video_seam_patches.py | 187 + tests/test_p24_comfy_integration.py | 179 + tests/test_p25_gpu_acceptance_probe.py | 59 + tests/test_p31_prompt_conditioning_cache.py | 149 + tests/test_p32_api_runner.py | 35 + tests/test_p32_gpu_profile_probe.py | 141 + tests/test_p32_profile_server_static.py | 32 + tests/test_p46a_reference_video_memory.py | 100 + tests/test_packed_layout_preflight.py | 269 ++ tests/test_preview_reuse_diagnostics.py | 162 + tests/test_preview_reuse_frontend.py | 13 + tests/test_public_api.py | 20 + tests/test_public_workflow_integrity.py | 61 + tests/test_release_metadata.py | 30 + tests/test_review_identity_lifecycle.py | 113 + tests/test_review_provenance_reroll.py | 50 + tests/test_review_rehydration.py | 20 + ...st_runtime_redesign_r0_characterization.py | 403 ++ ...st_runtime_redesign_r2_validated_prefix.py | 214 + .../test_runtime_redesign_r3_pure_planner.py | 220 + tests/test_runtime_redesign_r4_coordinator.py | 209 + .../test_runtime_redesign_r5_transactions.py | 318 ++ tests/test_runtime_redesign_r6_node_facade.py | 204 + tests/test_spectrum_portability.py | 104 + tests/test_state.py | 83 + tests/test_state_io.py | 49 + tests/test_storage_api.py | 80 + tests/test_storage_frontend.py | 14 + tests/test_temporal.py | 97 + tests/test_temporary_storage.py | 471 +++ tests/test_terminal_merge_diagnostics.py | 183 + tests/test_timeline_video.py | 215 + tests/test_v217_stable_facade.py | 148 + tests/test_v2_decoder.py | 87 + tests/test_v2_prompts.py | 541 +++ tests/test_v2_registration.py | 40 + tests/test_v2_seam.py | 127 + tests/test_v2_seeds_motion.py | 46 + tests/test_v2_sequence_orchestration.py | 99 + tests/test_v2_session.py | 112 + tests/test_v2_terminal_merge.py | 210 + tests/test_v2_widget_migration.py | 28 + tests/test_v31_run_storage.py | 1219 ++++++ tests/test_v32_reference.py | 186 + tests/test_v34_driving_audio.py | 302 ++ tests/test_v34_hybrid_reference.py | 142 + tests/test_v351_reference_audio.py | 154 + tests/test_v351_reference_audio_ui.py | 166 + tests/test_v35_compact_ui_frontend.py | 182 + tests/test_v35_conditioning_bridge.py | 249 ++ tests/test_v35_context_sampler.py | 533 +++ tests/test_v35_hires_fix.py | 567 +++ tests/test_v35_latent_resize.py | 152 + tests/test_v35_pixel_vae_resize.py | 341 ++ tests/test_v35_refine_context.py | 314 ++ tests/test_v35_refine_sampling.py | 112 + tests/test_v35_second_pass_contract.py | 145 + tests/test_v35_second_pass_sampling.py | 555 +++ tests/test_v36_chunk_duration.py | 45 + tests/test_v36_masked_video_prefix.py | 795 ++++ tests/test_v36_public_sampler.py | 185 + tests/test_v37_guide_production.py | 328 ++ tests/test_v37_guide_timeline.py | 151 + tests/test_v37_refine_schedule.py | 119 + .../test_v38_adaptive_continuity_observer.py | 253 ++ tests/test_v38_audio_only_refine.py | 292 ++ tests/test_v38_branch_provenance.py | 114 + tests/test_v38_chunk_boundary_diagnostics.py | 334 ++ tests/test_v38_chunk_scoped_refine.py | 437 ++ tests/test_v38_completed_review_ui.py | 11 + tests/test_v38_easy_mode.py | 592 +++ tests/test_v38_hover_help.py | 181 + tests/test_v38_memory_action_policy.py | 261 ++ tests/test_v38_multi_reference_audio.py | 322 ++ tests/test_v38_open_integration_contract.py | 248 ++ tests/test_v38_packed_row_planner.py | 206 + tests/test_v38_public_surface_cleanup.py | 698 +++ tests/test_v38_public_workflow.py | 269 ++ tests/test_v38_ref_encode_cache.py | 492 +++ tests/test_v38_refine_target.py | 312 ++ tests/test_v38_reliability.py | 529 +++ tests/test_v38_resolution_preset.py | 552 +++ tests/test_v38_review_cache.py | 319 ++ tests/test_v38_review_control.py | 451 ++ tests/test_v38_review_execution_cap.py | 675 +++ tests/test_v38_review_run_storage.py | 1662 ++++++++ tests/test_v38_review_ui.py | 481 +++ tests/test_v38_selective_second_pass_node.py | 368 ++ tests/test_v38_temporal_window_refine.py | 402 ++ tests/test_v38_video_audio_refine.py | 339 ++ tests/test_v38x2_decode_cache_integration.py | 60 + tests/test_v39_official_workflow.py | 106 + tests/test_v39_r0_fixed_prompt_cache.py | 156 + tests/test_v39_review_refine_scope.py | 157 + tests/test_v39_rr_r1_fixed_identity.py | 233 + tests/test_v39_rr_r2_reference_routing.py | 236 ++ tests/test_v39_rr_r3_effective_plan.py | 289 ++ tests/test_v39_rr_r4_reference_runtime.py | 250 ++ tests/test_v39_rr_r4_sequence_integration.py | 250 ++ ...st_v39_rr_r5_reference_storage_contract.py | 205 + tests/test_v39_rr_r6_public.py | 330 ++ tests/test_v39_stability_r1.py | 44 + tests/test_v39_stability_r2.py | 225 + tests/test_v39_stability_r3.py | 178 + tests/test_v39_stability_r4_r5.py | 139 + tests/test_v39_turbo_sample.py | 36 + tests/test_v3_hardening.py | 220 + tests/test_v3_physical_decode_groups.py | 153 + tests/test_v3_video_seam.py | 353 ++ tests/test_video_adapter.py | 49 + ...test_video_reference_modes_experimental.py | 242 ++ timeline_video.py | 342 ++ tools/__init__.py | 1 + tools/a2_audio_only_canary_node/__init__.py | 230 + .../__init__.py | 387 ++ tools/audio_continuity_output_analyze.py | 194 + tools/audio_continuity_research_runner.py | 163 + tools/g4_sampling_diagnostic_node/__init__.py | 3 + tools/g4_sampling_diagnostic_node/node.py | 646 +++ tools/gpu_diagnostic_node/README.md | 32 + tools/gpu_diagnostic_node/__init__.py | 9 + tools/gpu_diagnostic_node/capture.py | 302 ++ .../continuation_policy.py | 222 + tools/gpu_diagnostic_node/nodes.py | 1625 +++++++ tools/gpu_diagnostic_node/soft_reanchor.py | 164 + tools/install_gpu_diagnostic_node.ps1 | 29 + tools/issue13_drift_analyze.py | 180 + tools/issue13_drift_gate_runner.py | 262 ++ tools/issue13_r25_gate_runner.py | 408 ++ tools/issue13_r28_analyze.py | 292 ++ tools/issue13_r28_gate_runner.py | 258 ++ tools/issue13_r2_analyze.py | 462 ++ tools/issue13_r2_gate_runner.py | 282 ++ tools/p20_file_backed_tensor_poc.py | 1044 +++++ tools/p24_comfy_integration_probe.py | 676 +++ tools/p25_gpu_acceptance_probe.py | 963 +++++ tools/p32_api_runner.py | 189 + tools/p32_gpu_profile_probe.py | 719 ++++ tools/p32_profile_server.py | 300 ++ tools/snapshot.ps1 | 37 + tools/v36_fl2va_combined_gate_runner.py | 437 ++ tools/v36_pig3_reference_gate_runner.py | 167 + .../v36_production_integration_gate_runner.py | 273 ++ tools/v36_r1_api_runner.py | 134 + tools/v36_r1_stress_gate_analyze.py | 256 ++ tools/v36_r1_stress_gate_runner.py | 416 ++ tools/v36_r2_audio_boundary_analyze.py | 170 + tools/v36_r2_cached_seam_probe.py | 61 + tools/v36_r2_gpu_gate_runner.py | 292 ++ tools/v38_a2_audio_only_gpu_canary.py | 619 +++ tools/v38_a3c_selective_refine_gpu_matrix.py | 751 ++++ .../v38_a4_chunk_scoped_refine_gpu_matrix.py | 524 +++ ...38_a6_temporal_window_refine_gpu_matrix.py | 698 +++ tools/v38_a7a_packed_row_gpu_matrix.py | 632 +++ tools/v38_a7b_memory_action_gpu_matrix.py | 571 +++ .../v38_a8a_adaptive_continuity_gpu_matrix.py | 788 ++++ .../v38_easy_driving_audio_gpu_gate_runner.py | 397 ++ tools/v38_easy_gpu_gate_runner.py | 581 +++ tools/v38_g4_conditioning_determinism_gate.py | 345 ++ tools/v38_g4_sampling_determinism_gate.py | 719 ++++ tools/v38_resolution_gpu_gate_runner.py | 430 ++ tools/v38_review_phase_f_gpu_gate_runner.py | 415 ++ tools/v38_review_phase_f_recovery_runner.py | 197 + tools/validate.ps1 | 29 + tools/verify_runtime.py | 131 + v2/__init__.py | 5 + v2/context_diagnostics.py | 108 + v2/decoder.py | 124 + v2/diagnostics.py | 31 + v2/h3_builder.py | 362 ++ v2/motion.py | 58 + v2/nodes.py | 933 ++++ v2/prompts.py | 251 ++ v2/sampling.py | 124 + v2/seam_guard.py | 151 + v2/seam_types.py | 37 + v2/seeds.py | 36 + v2/sequence.py | 1806 ++++++++ v2/session.py | 265 ++ v2/session_io.py | 72 + v3/__init__.py | 5 + v3/adaptive_continuity.py | 320 ++ v3/assembly.py | 431 ++ v3/assembly_v35.py | 735 ++++ v3/conditioning_bridge_nodes.py | 135 + v3/driving_nodes.py | 1222 ++++++ v3/easy_audio_nodes.py | 67 + v3/easy_image_nodes.py | 48 + v3/easy_nodes.py | 482 +++ v3/easy_video_nodes.py | 274 ++ v3/execution_planner.py | 191 + v3/file_backed_buffer.py | 282 ++ v3/hires_fix_nodes.py | 527 +++ v3/latent_resize_nodes.py | 140 + v3/masked_continuation.py | 236 ++ v3/memory_action_policy.py | 221 + v3/memory_attribution.py | 416 ++ v3/node_facade.py | 176 + v3/nodes.py | 1350 ++++++ v3/packed_row_planner.py | 349 ++ v3/pixel_vae_resize.py | 367 ++ v3/plan.py | 466 ++ v3/planning_types.py | 83 + v3/ref_encode_cache.py | 340 ++ v3/reference_effective_plan.py | 287 ++ v3/reference_images_v39.py | 111 + v3/reference_plan_inspector.py | 69 + v3/reference_routing.py | 365 ++ v3/reference_runtime.py | 360 ++ v3/reference_storage_contract.py | 379 ++ v3/reference_video.py | 281 ++ v3/refine_context.py | 699 +++ v3/refine_sampling.py | 158 + v3/refine_schedule.py | 214 + v3/refine_scope.py | 179 + v3/refine_target.py | 206 + v3/refine_window.py | 421 ++ v3/reliability_v38.py | 920 ++++ v3/resolution.py | 248 ++ v3/review_control.py | 812 ++++ v3/runtime_coordinator.py | 223 + v3/sampling_engine.py | 71 + v3/second_pass.py | 644 +++ v3/second_pass_nodes.py | 385 ++ v3/targeted_refine_sampling.py | 496 +++ v3/targeted_second_pass.py | 491 +++ v3/video_adapter_nodes.py | 56 + v3/video_seam.py | 532 +++ v3/video_seam_v35.py | 131 + version.py | 8 + video_reference_modes.py | 299 ++ web/decode_cache_helper.js | 105 + web/easy_bypass_toggle.js | 111 + web/easy_load_audio.js | 69 + web/easy_load_image.js | 29 + web/easy_load_video.js | 3 + web/easy_v38.js | 135 + web/project_id.js | 3562 ++++++++++++++++ web/reference_audio_ui.js | 29 + web/reference_image_ui.js | 71 + web/reference_images_v39.js | 264 ++ web/storage_settings.js | 240 ++ web/video_reference_mode_ui.js | 42 + write-h3-v39-reference-prompts.zip | Bin 0 -> 5341 bytes write-h3-v39-reference-prompts/SKILL.md | 51 + .../references/simple-3x5.txt | 14 + .../references/structured.md | 36 + 481 files changed, 131517 insertions(+) create mode 100644 .comfyignore create mode 100644 .gitattributes create mode 100644 .github/workflows/publish_registry.yml create mode 100644 .github/workflows/publish_release.yml create mode 100644 .github/workflows/tests.yml create mode 100644 .gitignore create mode 100644 AGENTS.md create mode 100644 CHANGELOG.md create mode 100644 H3-Continuum-LLM-System-Prompt-v1.zip create mode 100644 H3-Continuum-Skill-v1.zip create mode 100644 INSTALL_JA.txt create mode 100644 LICENSE create mode 100644 MANIFEST.sha256 create mode 100644 PACKAGE_INFO_JA.txt create mode 100644 PACKAGE_VALIDATION.txt create mode 100644 PROJECT_STATE.md create mode 100644 README.md create mode 100644 README_JA.md create mode 100644 REGISTRY_MANIFEST.sha256 create mode 100644 THIRD_PARTY_NOTICES.md create mode 100644 WORKLOG.md create mode 100644 __init__.py create mode 100644 audio_compat.py create mode 100644 branch_provenance.py create mode 100644 compatibility.py create mode 100644 conditioning.py create mode 100644 constants.py create mode 100644 continuation.py create mode 100644 decode_cache/__init__.py create mode 100644 decode_cache/identity.py create mode 100644 decode_cache/nodes.py create mode 100644 decode_cache/service.py create mode 100644 decode_cache/store.py create mode 100644 docs/AB_TEST_LOG_JA.md create mode 100644 docs/ARCHITECTURE.md create mode 100644 docs/FAILURE_SEMANTICS_BASELINE.md create mode 100644 docs/FAILURE_SEMANTICS_DELTA_R5.md create mode 100644 docs/FUTURE_ROADMAP.md create mode 100644 docs/ISSUE31_SAMPLING_SIGNATURES.md create mode 100644 docs/KJ_PREVIEW_OVERRIDE_V398.md create mode 100644 docs/LOADER_PERSISTENCE_REPAIR.md create mode 100644 docs/N2B_BRANCH_PROVENANCE_CONTRACT_V1.md create mode 100644 docs/R0_RUNTIME_CHARACTERIZATION.md create mode 100644 docs/R6_NODE_FACADE_DELTA.md create mode 100644 docs/README_HISTORY.md create mode 100644 docs/README_HISTORY_JA.md create mode 100644 docs/REGENERATE_INTENT_V396.md create mode 100644 docs/RELEASE_NOTES_V3910.md create mode 100644 docs/RELEASE_NOTES_V395.md create mode 100644 docs/RELEASE_NOTES_V396.md create mode 100644 docs/RELEASE_NOTES_V397.md create mode 100644 docs/RELEASE_NOTES_V398.md create mode 100644 docs/RELEASE_NOTES_V399.md create mode 100644 docs/REVIEW_RESTORATION_REPAIR.md create mode 100644 docs/SEAM_TEST_SOURCE_JA.md create mode 100644 docs/TEMPORARY_STORAGE_JA.md create mode 100644 docs/TIMELINE_VIDEO_EXPERIMENTAL.md create mode 100644 docs/TURBO_SAMPLE.md create mode 100644 docs/USER_GUIDE.md create mode 100644 docs/USER_GUIDE_JA.md create mode 100644 docs/V2_1_4_UI_SIMPLIFICATION.md create mode 100644 docs/V2_1_5_UI_SERIALIZATION_STABILITY.md create mode 100644 docs/V2_1_6_UNIFIED_UI.md create mode 100644 docs/V34_DRIVING_AUDIO_VALIDATION.md create mode 100644 docs/V35_HIRES_FIX.md create mode 100644 docs/V38_BEGINNER_MANUAL_PLAN_AND_CAPTURE_SCRIPT_JA.md create mode 100644 docs/V38_OPEN_INTEGRATION_CONTRACT.md create mode 100644 docs/V38_RELEASE_AND_MIGRATION.md create mode 100644 docs/V39_REFERENCE_INLINE.md create mode 100644 docs/VALIDATION_JA.md create mode 100644 docs/VALIDATION_RESULTS.md create mode 100644 docs/images/comfyui-manager-h3-continuum.png create mode 100644 docs/images/h3-continuum-advanced.png create mode 100644 docs/images/h3-continuum-detailed-report.png create mode 100644 docs/images/h3-continuum-image-scale.png create mode 100644 docs/images/h3-continuum-sampler.png create mode 100644 docs/images/h3-continuum-v33-settings.png create mode 100644 docs/images/h3-continuum-v33-streamlined-ui.png create mode 100644 docs/images/v31b-auto-resume.png create mode 100644 docs/images/v34-driving-audio-connection.png create mode 100644 docs/images/v34-regenerate-from.png create mode 100644 docs/images/v34-sampler-spectrum-assemble.png create mode 100644 docs/images/v34-video-reference-separate-audio.png create mode 100644 docs/images/v34-video-reference-with-audio.png create mode 100644 docs/images/v34-workflow-overview.png create mode 100644 docs/images/v35-assemble-seam-node.png create mode 100644 docs/images/v35-hires-fix-node.png create mode 100644 docs/images/v35-second-pass-node.png create mode 100644 docs/images/v351-lbh-conditioning-bridge-flow.svg create mode 100644 docs/images/v351-reference-audio-permanent.png create mode 100644 docs/images/v351-video-guide-frames.png create mode 100644 docs/images/v352-stabilization-optimization.png create mode 100644 docs/images/v36-masked-av-continuation.png create mode 100644 docs/images/v36-sampler-node.png create mode 100644 docs/images/v37-sampler-node.png create mode 100644 docs/images/v37-tail6-second-pass.png create mode 100644 docs/images/v38-manual/continue-selected.png create mode 100644 docs/images/v38-manual/first-last-image-bypass.png create mode 100644 docs/images/v38-manual/full-video-progress.png create mode 100644 docs/images/v38-manual/h3-continuum-v38-civitai.png create mode 100644 docs/images/v38-manual/h3-continuum-v38-whats-new.png create mode 100644 docs/images/v38-manual/input-video-audio-bypass.png create mode 100644 docs/images/v38-manual/live-continue-selected-steps4.png create mode 100644 docs/images/v38-manual/live-execute-button.png create mode 100644 docs/images/v38-manual/live-first-review.png create mode 100644 docs/images/v38-manual/live-regenerate-from-chunk2.png create mode 100644 docs/images/v38-manual/live-render-history-steps4.png create mode 100644 docs/images/v38-manual/live-retry-selected.png create mode 100644 docs/images/v38-manual/live-review-actions-steps4.png create mode 100644 docs/images/v38-manual/live-review-settings.png create mode 100644 docs/images/v38-manual/live-sequence-complete-steps4.png create mode 100644 docs/images/v38-manual/live-settings-complete-steps4.png create mode 100644 docs/images/v38-manual/reference-images-three-bypass.png create mode 100644 docs/images/v38-manual/return-to-review.png create mode 100644 docs/images/v38-manual/review-actions.png create mode 100644 docs/images/v38-manual/review-settings.png create mode 100644 docs/images/v38-manual/sampler-size-first-image-draft.png create mode 100644 docs/images/v38-manual/seed-fixed.png create mode 100644 docs/images/v38-manual/size-source-first-image-tooltip.png create mode 100644 docs/images/v38-manual/size-source-manual-640.png create mode 100644 docs/images/v38-manual/turbo-lora-choices.png create mode 100644 docs/images/v39/v38x2-feature-summary-historical.png create mode 100644 docs/images/v39/v39-feature-summary.png create mode 100644 docs/images/v39/v39-reference-images-node.png create mode 100644 docs/images/v39/v39-sampler-node.png create mode 100644 driving_audio.py create mode 100644 examples/H3_CONTINUUM_V1_WIRING.md create mode 100644 examples/H3_CONTINUUM_V2_WIRING_JA.md create mode 100644 examples/H3_Continuum_V2_README.txt create mode 100644 examples/H3_Continuum_V2_reference.webp create mode 100644 examples/MiniMax_H3_Continuum_V38X2.json create mode 100644 examples/MiniMax_H3_Continuum_V39.json create mode 100644 examples/MiniMax_H3_Continuum_V39_Turbo_Sample.json create mode 100644 examples/settings/recommended_settings_v1.json create mode 100644 examples/settings/recommended_settings_v2.json create mode 100644 examples/v2_prompt_list_example.txt create mode 100644 examples/v2_prompt_timeline_example.txt create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.json create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.zip create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38X2.json create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38X2.zip create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Follow.json create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Follow.zip create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Repeat.json create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Repeat.zip create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38x.json create mode 100644 examples/workflows/MiniMax_H3_Continuum_V38x.zip create mode 100644 examples/workflows/MiniMax_H3_Continuum_V39.json create mode 100644 examples/workflows/MiniMax_H3_Continuum_V39.zip create mode 100644 examples/workflows/MiniMax_H3_Continuum_V39_Reference_Inline.json create mode 100644 examples/workflows/MiniMax_H3_Continuum_V39_Reference_Inline.zip create mode 100644 examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.json create mode 100644 examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.zip create mode 100644 graph_contract.py create mode 100644 guide_timeline.py create mode 100644 hardening.py create mode 100644 install_windows.bat create mode 100644 install_windows.ps1 create mode 100644 layout_adapter.py create mode 100644 media.py create mode 100644 metadata.ini create mode 100644 minimax-h3-continuum-dance-director.zip create mode 100644 model_patch.py create mode 100644 nodes.py create mode 100644 preview_override_provider_public_v1.json create mode 100644 preview_override_provider_v1.json create mode 100644 preview_override_signature.py create mode 100644 public_api.py create mode 100644 pyproject.toml create mode 100644 pytest.ini create mode 100644 reference.py create mode 100644 reference_audio.py create mode 100644 reference_video.py create mode 100644 requirements.txt create mode 100644 run_storage.py create mode 100644 spectrum_provider_v029.json create mode 100644 spectrum_provider_v029_py312.json create mode 100644 spectrum_provider_v1.json create mode 100644 spectrum_provider_v1_py312.json create mode 100644 spectrum_signature.py create mode 100644 state.py create mode 100644 state_io.py create mode 100644 storage_api.py create mode 100644 storage_fs.py create mode 100644 storage_memory.py create mode 100644 storage_settings.py create mode 100644 temporal.py create mode 100644 tests/conftest.py create mode 100644 tests/decode_cache/conftest.py create mode 100644 tests/decode_cache/helpers/cache_process_worker.py create mode 100644 tests/decode_cache/test_adapter_nodes.py create mode 100644 tests/decode_cache/test_identity.py create mode 100644 tests/decode_cache/test_process_isolation.py create mode 100644 tests/decode_cache/test_store.py create mode 100644 tests/fixtures/issue31_local_spectrum_signature.py.txt create mode 100644 tests/fixtures/issue31_phase1_model_options.py.txt create mode 100644 tests/fixtures/v38_a3c_selective_refine_saved_workflow.json create mode 100644 tests/fixtures/v39_sampling_v5/project.json create mode 100644 tests/fixtures/v39_sampling_v5/revisions/d8352287737286e5/chunks/rev-d8352287737286e5-txn-008f52b93a88461481ead25f91627af0-chunk-0002.safetensors create mode 100644 tests/fixtures/v39_sampling_v5/revisions/d8352287737286e5/chunks/rev-d8352287737286e5-txn-9bb0a584b7344342b114bd31b175225c-chunk-0001.safetensors create mode 100644 tests/fixtures/v39_sampling_v5/revisions/d8352287737286e5/manifest.json create mode 100644 tests/frontend_decode_cache_helper.mjs create mode 100644 tests/frontend_external_int_size.cjs create mode 100644 tests/frontend_issue30_ui.cjs create mode 100644 tests/frontend_reference_images_v39.cjs create mode 100644 tests/frontend_review_queue.cjs create mode 100644 tests/frontend_storage_settings.mjs create mode 100644 tests/helpers/r5_process_crash_worker.py create mode 100644 tests/terminal_merge_diagnostics.py create mode 100644 tests/test_audit_boundary_repairs.py create mode 100644 tests/test_compatibility.py create mode 100644 tests/test_conditioning_modes.py create mode 100644 tests/test_context_diagnostics.py create mode 100644 tests/test_easy_load_audio.py create mode 100644 tests/test_easy_load_image.py create mode 100644 tests/test_easy_load_video.py create mode 100644 tests/test_external_int_size.py create mode 100644 tests/test_gpu_diagnostic_runner.py create mode 100644 tests/test_issue13_r25_tools.py create mode 100644 tests/test_issue13_r28_tools.py create mode 100644 tests/test_issue13_r2_tools.py create mode 100644 tests/test_issue30_frontend.py create mode 100644 tests/test_issue31_attention_signature.py create mode 100644 tests/test_issue31_preview_override_signature.py create mode 100644 tests/test_issue31_spectrum_signature.py create mode 100644 tests/test_layout_adapter.py create mode 100644 tests/test_loader_store_roundtrip.py create mode 100644 tests/test_m3_conditioning_display.py create mode 100644 tests/test_model_patch.py create mode 100644 tests/test_nodes_logic.py create mode 100644 tests/test_p20_file_backed_tensor_poc.py create mode 100644 tests/test_p21_memory_attribution.py create mode 100644 tests/test_p22_file_backed_buffer.py create mode 100644 tests/test_p23_assembly_v35.py create mode 100644 tests/test_p23_video_seam_patches.py create mode 100644 tests/test_p24_comfy_integration.py create mode 100644 tests/test_p25_gpu_acceptance_probe.py create mode 100644 tests/test_p31_prompt_conditioning_cache.py create mode 100644 tests/test_p32_api_runner.py create mode 100644 tests/test_p32_gpu_profile_probe.py create mode 100644 tests/test_p32_profile_server_static.py create mode 100644 tests/test_p46a_reference_video_memory.py create mode 100644 tests/test_packed_layout_preflight.py create mode 100644 tests/test_preview_reuse_diagnostics.py create mode 100644 tests/test_preview_reuse_frontend.py create mode 100644 tests/test_public_api.py create mode 100644 tests/test_public_workflow_integrity.py create mode 100644 tests/test_release_metadata.py create mode 100644 tests/test_review_identity_lifecycle.py create mode 100644 tests/test_review_provenance_reroll.py create mode 100644 tests/test_review_rehydration.py create mode 100644 tests/test_runtime_redesign_r0_characterization.py create mode 100644 tests/test_runtime_redesign_r2_validated_prefix.py create mode 100644 tests/test_runtime_redesign_r3_pure_planner.py create mode 100644 tests/test_runtime_redesign_r4_coordinator.py create mode 100644 tests/test_runtime_redesign_r5_transactions.py create mode 100644 tests/test_runtime_redesign_r6_node_facade.py create mode 100644 tests/test_spectrum_portability.py create mode 100644 tests/test_state.py create mode 100644 tests/test_state_io.py create mode 100644 tests/test_storage_api.py create mode 100644 tests/test_storage_frontend.py create mode 100644 tests/test_temporal.py create mode 100644 tests/test_temporary_storage.py create mode 100644 tests/test_terminal_merge_diagnostics.py create mode 100644 tests/test_timeline_video.py create mode 100644 tests/test_v217_stable_facade.py create mode 100644 tests/test_v2_decoder.py create mode 100644 tests/test_v2_prompts.py create mode 100644 tests/test_v2_registration.py create mode 100644 tests/test_v2_seam.py create mode 100644 tests/test_v2_seeds_motion.py create mode 100644 tests/test_v2_sequence_orchestration.py create mode 100644 tests/test_v2_session.py create mode 100644 tests/test_v2_terminal_merge.py create mode 100644 tests/test_v2_widget_migration.py create mode 100644 tests/test_v31_run_storage.py create mode 100644 tests/test_v32_reference.py create mode 100644 tests/test_v34_driving_audio.py create mode 100644 tests/test_v34_hybrid_reference.py create mode 100644 tests/test_v351_reference_audio.py create mode 100644 tests/test_v351_reference_audio_ui.py create mode 100644 tests/test_v35_compact_ui_frontend.py create mode 100644 tests/test_v35_conditioning_bridge.py create mode 100644 tests/test_v35_context_sampler.py create mode 100644 tests/test_v35_hires_fix.py create mode 100644 tests/test_v35_latent_resize.py create mode 100644 tests/test_v35_pixel_vae_resize.py create mode 100644 tests/test_v35_refine_context.py create mode 100644 tests/test_v35_refine_sampling.py create mode 100644 tests/test_v35_second_pass_contract.py create mode 100644 tests/test_v35_second_pass_sampling.py create mode 100644 tests/test_v36_chunk_duration.py create mode 100644 tests/test_v36_masked_video_prefix.py create mode 100644 tests/test_v36_public_sampler.py create mode 100644 tests/test_v37_guide_production.py create mode 100644 tests/test_v37_guide_timeline.py create mode 100644 tests/test_v37_refine_schedule.py create mode 100644 tests/test_v38_adaptive_continuity_observer.py create mode 100644 tests/test_v38_audio_only_refine.py create mode 100644 tests/test_v38_branch_provenance.py create mode 100644 tests/test_v38_chunk_boundary_diagnostics.py create mode 100644 tests/test_v38_chunk_scoped_refine.py create mode 100644 tests/test_v38_completed_review_ui.py create mode 100644 tests/test_v38_easy_mode.py create mode 100644 tests/test_v38_hover_help.py create mode 100644 tests/test_v38_memory_action_policy.py create mode 100644 tests/test_v38_multi_reference_audio.py create mode 100644 tests/test_v38_open_integration_contract.py create mode 100644 tests/test_v38_packed_row_planner.py create mode 100644 tests/test_v38_public_surface_cleanup.py create mode 100644 tests/test_v38_public_workflow.py create mode 100644 tests/test_v38_ref_encode_cache.py create mode 100644 tests/test_v38_refine_target.py create mode 100644 tests/test_v38_reliability.py create mode 100644 tests/test_v38_resolution_preset.py create mode 100644 tests/test_v38_review_cache.py create mode 100644 tests/test_v38_review_control.py create mode 100644 tests/test_v38_review_execution_cap.py create mode 100644 tests/test_v38_review_run_storage.py create mode 100644 tests/test_v38_review_ui.py create mode 100644 tests/test_v38_selective_second_pass_node.py create mode 100644 tests/test_v38_temporal_window_refine.py create mode 100644 tests/test_v38_video_audio_refine.py create mode 100644 tests/test_v38x2_decode_cache_integration.py create mode 100644 tests/test_v39_official_workflow.py create mode 100644 tests/test_v39_r0_fixed_prompt_cache.py create mode 100644 tests/test_v39_review_refine_scope.py create mode 100644 tests/test_v39_rr_r1_fixed_identity.py create mode 100644 tests/test_v39_rr_r2_reference_routing.py create mode 100644 tests/test_v39_rr_r3_effective_plan.py create mode 100644 tests/test_v39_rr_r4_reference_runtime.py create mode 100644 tests/test_v39_rr_r4_sequence_integration.py create mode 100644 tests/test_v39_rr_r5_reference_storage_contract.py create mode 100644 tests/test_v39_rr_r6_public.py create mode 100644 tests/test_v39_stability_r1.py create mode 100644 tests/test_v39_stability_r2.py create mode 100644 tests/test_v39_stability_r3.py create mode 100644 tests/test_v39_stability_r4_r5.py create mode 100644 tests/test_v39_turbo_sample.py create mode 100644 tests/test_v3_hardening.py create mode 100644 tests/test_v3_physical_decode_groups.py create mode 100644 tests/test_v3_video_seam.py create mode 100644 tests/test_video_adapter.py create mode 100644 tests/test_video_reference_modes_experimental.py create mode 100644 timeline_video.py create mode 100644 tools/__init__.py create mode 100644 tools/a2_audio_only_canary_node/__init__.py create mode 100644 tools/a3c_selective_refine_probe_node/__init__.py create mode 100644 tools/audio_continuity_output_analyze.py create mode 100644 tools/audio_continuity_research_runner.py create mode 100644 tools/g4_sampling_diagnostic_node/__init__.py create mode 100644 tools/g4_sampling_diagnostic_node/node.py create mode 100644 tools/gpu_diagnostic_node/README.md create mode 100644 tools/gpu_diagnostic_node/__init__.py create mode 100644 tools/gpu_diagnostic_node/capture.py create mode 100644 tools/gpu_diagnostic_node/continuation_policy.py create mode 100644 tools/gpu_diagnostic_node/nodes.py create mode 100644 tools/gpu_diagnostic_node/soft_reanchor.py create mode 100644 tools/install_gpu_diagnostic_node.ps1 create mode 100644 tools/issue13_drift_analyze.py create mode 100644 tools/issue13_drift_gate_runner.py create mode 100644 tools/issue13_r25_gate_runner.py create mode 100644 tools/issue13_r28_analyze.py create mode 100644 tools/issue13_r28_gate_runner.py create mode 100644 tools/issue13_r2_analyze.py create mode 100644 tools/issue13_r2_gate_runner.py create mode 100644 tools/p20_file_backed_tensor_poc.py create mode 100644 tools/p24_comfy_integration_probe.py create mode 100644 tools/p25_gpu_acceptance_probe.py create mode 100644 tools/p32_api_runner.py create mode 100644 tools/p32_gpu_profile_probe.py create mode 100644 tools/p32_profile_server.py create mode 100644 tools/snapshot.ps1 create mode 100644 tools/v36_fl2va_combined_gate_runner.py create mode 100644 tools/v36_pig3_reference_gate_runner.py create mode 100644 tools/v36_production_integration_gate_runner.py create mode 100644 tools/v36_r1_api_runner.py create mode 100644 tools/v36_r1_stress_gate_analyze.py create mode 100644 tools/v36_r1_stress_gate_runner.py create mode 100644 tools/v36_r2_audio_boundary_analyze.py create mode 100644 tools/v36_r2_cached_seam_probe.py create mode 100644 tools/v36_r2_gpu_gate_runner.py create mode 100644 tools/v38_a2_audio_only_gpu_canary.py create mode 100644 tools/v38_a3c_selective_refine_gpu_matrix.py create mode 100644 tools/v38_a4_chunk_scoped_refine_gpu_matrix.py create mode 100644 tools/v38_a6_temporal_window_refine_gpu_matrix.py create mode 100644 tools/v38_a7a_packed_row_gpu_matrix.py create mode 100644 tools/v38_a7b_memory_action_gpu_matrix.py create mode 100644 tools/v38_a8a_adaptive_continuity_gpu_matrix.py create mode 100644 tools/v38_easy_driving_audio_gpu_gate_runner.py create mode 100644 tools/v38_easy_gpu_gate_runner.py create mode 100644 tools/v38_g4_conditioning_determinism_gate.py create mode 100644 tools/v38_g4_sampling_determinism_gate.py create mode 100644 tools/v38_resolution_gpu_gate_runner.py create mode 100644 tools/v38_review_phase_f_gpu_gate_runner.py create mode 100644 tools/v38_review_phase_f_recovery_runner.py create mode 100644 tools/validate.ps1 create mode 100644 tools/verify_runtime.py create mode 100644 v2/__init__.py create mode 100644 v2/context_diagnostics.py create mode 100644 v2/decoder.py create mode 100644 v2/diagnostics.py create mode 100644 v2/h3_builder.py create mode 100644 v2/motion.py create mode 100644 v2/nodes.py create mode 100644 v2/prompts.py create mode 100644 v2/sampling.py create mode 100644 v2/seam_guard.py create mode 100644 v2/seam_types.py create mode 100644 v2/seeds.py create mode 100644 v2/sequence.py create mode 100644 v2/session.py create mode 100644 v2/session_io.py create mode 100644 v3/__init__.py create mode 100644 v3/adaptive_continuity.py create mode 100644 v3/assembly.py create mode 100644 v3/assembly_v35.py create mode 100644 v3/conditioning_bridge_nodes.py create mode 100644 v3/driving_nodes.py create mode 100644 v3/easy_audio_nodes.py create mode 100644 v3/easy_image_nodes.py create mode 100644 v3/easy_nodes.py create mode 100644 v3/easy_video_nodes.py create mode 100644 v3/execution_planner.py create mode 100644 v3/file_backed_buffer.py create mode 100644 v3/hires_fix_nodes.py create mode 100644 v3/latent_resize_nodes.py create mode 100644 v3/masked_continuation.py create mode 100644 v3/memory_action_policy.py create mode 100644 v3/memory_attribution.py create mode 100644 v3/node_facade.py create mode 100644 v3/nodes.py create mode 100644 v3/packed_row_planner.py create mode 100644 v3/pixel_vae_resize.py create mode 100644 v3/plan.py create mode 100644 v3/planning_types.py create mode 100644 v3/ref_encode_cache.py create mode 100644 v3/reference_effective_plan.py create mode 100644 v3/reference_images_v39.py create mode 100644 v3/reference_plan_inspector.py create mode 100644 v3/reference_routing.py create mode 100644 v3/reference_runtime.py create mode 100644 v3/reference_storage_contract.py create mode 100644 v3/reference_video.py create mode 100644 v3/refine_context.py create mode 100644 v3/refine_sampling.py create mode 100644 v3/refine_schedule.py create mode 100644 v3/refine_scope.py create mode 100644 v3/refine_target.py create mode 100644 v3/refine_window.py create mode 100644 v3/reliability_v38.py create mode 100644 v3/resolution.py create mode 100644 v3/review_control.py create mode 100644 v3/runtime_coordinator.py create mode 100644 v3/sampling_engine.py create mode 100644 v3/second_pass.py create mode 100644 v3/second_pass_nodes.py create mode 100644 v3/targeted_refine_sampling.py create mode 100644 v3/targeted_second_pass.py create mode 100644 v3/video_adapter_nodes.py create mode 100644 v3/video_seam.py create mode 100644 v3/video_seam_v35.py create mode 100644 version.py create mode 100644 video_reference_modes.py create mode 100644 web/decode_cache_helper.js create mode 100644 web/easy_bypass_toggle.js create mode 100644 web/easy_load_audio.js create mode 100644 web/easy_load_image.js create mode 100644 web/easy_load_video.js create mode 100644 web/easy_v38.js create mode 100644 web/project_id.js create mode 100644 web/reference_audio_ui.js create mode 100644 web/reference_image_ui.js create mode 100644 web/reference_images_v39.js create mode 100644 web/storage_settings.js create mode 100644 web/video_reference_mode_ui.js create mode 100644 write-h3-v39-reference-prompts.zip create mode 100644 write-h3-v39-reference-prompts/SKILL.md create mode 100644 write-h3-v39-reference-prompts/references/simple-3x5.txt create mode 100644 write-h3-v39-reference-prompts/references/structured.md diff --git a/.comfyignore b/.comfyignore new file mode 100644 index 0000000..8d3a535 --- /dev/null +++ b/.comfyignore @@ -0,0 +1,99 @@ +tests/ +tools/* +!tools/__init__.py +!tools/p20_file_backed_tensor_poc.py +!tools/verify_runtime.py +.pytest_cache/ +**/__pycache__/ +*.pyc +*.pyo +# Local backup archives are not distribution assets. The official workflow +# archive is re-included explicitly below. +*.zip + +# Internal source/development records and the source-tree Manifest are not part +# of the Registry archive. REGISTRY_MANIFEST.sha256 covers the staged package. +AGENTS.md +PROJECT_STATE.md +WORKLOG.md +MANIFEST.sha256 + +# Repository-only automation and test configuration are not runtime assets. +.github/ +.gitignore +.gitattributes +pytest.ini + +# Ship current V3.9 and retained V3.8X2 public contracts. Historical validation notes, +# version-specific design documents, and their screenshots stay in source. +docs/* +!docs/N2B_BRANCH_PROVENANCE_CONTRACT_V1.md +!docs/V38_OPEN_INTEGRATION_CONTRACT.md +!docs/V38_RELEASE_AND_MIGRATION.md + +# README launch illustrations: include only the current V3.9 guide assets. +!docs/images/ +docs/images/* +!docs/images/v39/ +docs/images/v39/* +!docs/images/v39/v38x2-feature-summary-historical.png +!docs/images/v39/v39-feature-summary.png +!docs/images/v39/v39-reference-images-node.png +!docs/images/v39/v39-sampler-node.png + +# Ship the separate V3.9 and V3.8X2 workflows. The V3.8X2 standard ZIP also +# contains V3.9; its helper-explicit ZIP remains the V3.8X2-only alias. +# External custom-node dependencies are documented, not bundled. +examples/* +!examples/workflows/ +examples/workflows/* +!examples/workflows/MiniMax_H3_Continuum_V38X2.json +!examples/workflows/MiniMax_H3_Continuum_V38X2.zip +!examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.json +!examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.zip +!examples/workflows/MiniMax_H3_Continuum_V39_Reference_Inline.json +!examples/workflows/MiniMax_H3_Continuum_V39_Reference_Inline.zip +!examples/workflows/MiniMax_H3_Continuum_V39.json +!examples/workflows/MiniMax_H3_Continuum_V39.zip +!examples/MiniMax_H3_Continuum_V39.json +!examples/MiniMax_H3_Continuum_V39_Turbo_Sample.json +!examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.json +!examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.zip +!docs/TURBO_SAMPLE.md +!examples/MiniMax_H3_Continuum_V38X2.json + +# Current loader migration instructions. +!docs/LOADER_PERSISTENCE_REPAIR.md +!docs/TIMELINE_VIDEO_EXPERIMENTAL.md +!docs/V39_REFERENCE_INLINE.md +.h3-timeline-experimental.json + +# Current temporary storage controls and lifecycle contract. +!docs/TEMPORARY_STORAGE_JA.md + +# Keep the Registry payload aligned with its declared Manifest. +!docs/REVIEW_RESTORATION_REPAIR.md +write-h3-v39-reference-prompts/ + +# Current Issue #31 compatibility and update guidance. +!docs/ISSUE31_SAMPLING_SIGNATURES.md + +# Current regeneration intent and automatic variation guide. +!docs/REGENERATE_INTENT_V396.md + +# Current audited KJ Preview Override compatibility guide. +!docs/KJ_PREVIEW_OVERRIDE_V398.md + +# Keep the condensed README and its moved text references together. +!docs/USER_GUIDE.md +!docs/README_HISTORY.md + +# Japanese detailed references retained by the condensed README. +!docs/USER_GUIDE_JA.md +!docs/README_HISTORY_JA.md + +# Current 3.9.9 update and verification notes. +!docs/RELEASE_NOTES_V399.md + +# Current 3.9.10 repair/update notes. +!docs/RELEASE_NOTES_V3910.md diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..c297181 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,3 @@ +# Release manifests and the user-supplied workflow ZIP hash exact file bytes. +# Keep Git checkout/add from rewriting line endings on Windows or Linux. +* -text diff --git a/.github/workflows/publish_registry.yml b/.github/workflows/publish_registry.yml new file mode 100644 index 0000000..38a78fc --- /dev/null +++ b/.github/workflows/publish_registry.yml @@ -0,0 +1,79 @@ +name: Publish to Comfy Registry + +on: + workflow_dispatch: + +permissions: + contents: read + +concurrency: + group: publish-comfy-registry-v3.9.10 + cancel-in-progress: false + +jobs: + publish: + runs-on: ubuntu-latest + timeout-minutes: 15 + steps: + - name: Check out v3.9.10 release + uses: actions/checkout@v4 + with: + ref: v3.9.10 + persist-credentials: false + - name: Set up Python + uses: actions/setup-python@v5 + with: + python-version: "3.11" + - name: Verify release identity and source hashes + shell: bash + run: | + test "$(git describe --tags --exact-match HEAD)" = "v3.9.10" + python - <<'PY' + import hashlib + import pathlib + import tomllib + root = pathlib.Path(".") + metadata = tomllib.loads((root / "pyproject.toml").read_text()) + assert metadata["project"]["version"] == "3.9.10" + assert metadata["tool"]["comfy"]["PublisherId"] == "ukr8b3g-cmyk" + for name in ("REGISTRY_MANIFEST.sha256", "MANIFEST.sha256"): + rows = (root / name).read_text().splitlines() + for row in rows: + expected, path = row.split(" ", 1) + assert hashlib.sha256((root / path).read_bytes()).hexdigest() == expected, path + print(name, len(rows), "matched") + print("Release commit:", __import__("subprocess").check_output(["git", "rev-parse", "HEAD"], text=True).strip()) + PY + - name: Install publishing CLI + run: python -m pip install comfy-cli==1.22.0 + - name: Build Registry package + shell: bash + run: | + comfy --skip-prompt --no-enable-telemetry env + comfy node pack + - name: Verify actual Registry ZIP before publication + shell: bash + run: | + python - <<'PY' + import hashlib + import pathlib + import zipfile + rows = pathlib.Path("REGISTRY_MANIFEST.sha256").read_text().splitlines() + declared = dict(row.split(" ", 1)[::-1] for row in rows) + with zipfile.ZipFile("node.zip") as archive: + assert archive.testzip() is None, "ZIP CRC failure" + actual = {entry.filename for entry in archive.infolist() if not entry.is_dir()} + expected = set(declared) | {"REGISTRY_MANIFEST.sha256"} + assert actual == expected, {"missing": sorted(expected - actual), "extra": sorted(actual - expected)} + for path, digest in declared.items(): + assert hashlib.sha256(archive.read(path)).hexdigest() == digest, path + assert archive.read("REGISTRY_MANIFEST.sha256") == pathlib.Path("REGISTRY_MANIFEST.sha256").read_bytes() + print("Registry ZIP: exact file set and", len(declared), "hashes matched") + PY + - name: Publish verified package + env: + REGISTRY_ACCESS_TOKEN: ${{ secrets.REGISTRY_ACCESS_TOKEN }} + shell: bash + run: | + test -n "$REGISTRY_ACCESS_TOKEN" + comfy node publish --token "$REGISTRY_ACCESS_TOKEN" \ No newline at end of file diff --git a/.github/workflows/publish_release.yml b/.github/workflows/publish_release.yml new file mode 100644 index 0000000..23c1f1c --- /dev/null +++ b/.github/workflows/publish_release.yml @@ -0,0 +1,154 @@ +name: Publish verified v3.9.10 + +on: + workflow_run: + workflows: [tests] + types: [completed] + +permissions: + contents: read + +concurrency: + group: publish-v3.9.10 + cancel-in-progress: false + +jobs: + tag: + if: github.event.workflow_run.conclusion == 'success' && github.event.workflow_run.head_branch == 'main' && github.event.workflow_run.head_repository.full_name == github.repository + runs-on: ubuntu-latest + permissions: + contents: write + outputs: + publish: ${{ steps.tag.outputs.publish }} + steps: + - uses: actions/checkout@v4 + with: + ref: ${{ github.event.workflow_run.head_sha }} + - id: tag + env: + GH_TOKEN: ${{ github.token }} + RELEASE_SHA: ${{ github.event.workflow_run.head_sha }} + run: | + test "$(python -c 'import tomllib;print(tomllib.load(open("pyproject.toml","rb"))["project"]["version"])')" = 3.9.10 + test "$(gh api repos/$GITHUB_REPOSITORY/branches/main --jq .commit.sha)" = "$RELEASE_SHA" + if gh release view v3.9.10 --json isDraft --jq .isDraft 2>/dev/null | grep -qx false; then + echo 'publish=false' >> "$GITHUB_OUTPUT" + exit 0 + fi + if EXISTING=$(gh api repos/$GITHUB_REPOSITORY/git/ref/tags/v3.9.10 --jq .object.sha 2>/dev/null); then + echo "Existing tag: $EXISTING" + test "$EXISTING" = "$RELEASE_SHA" + else + echo "Creating v3.9.10 at $RELEASE_SHA" + gh api repos/$GITHUB_REPOSITORY/git/refs -f ref=refs/tags/v3.9.10 -f sha="$RELEASE_SHA" + fi + echo 'publish=true' >> "$GITHUB_OUTPUT" + tag-tests: + needs: tag + if: needs.tag.outputs.publish == 'true' + strategy: + fail-fast: false + matrix: + os: [ubuntu-latest, windows-latest] + runs-on: ${{ matrix.os }} + steps: + - uses: actions/checkout@v4 + with: + ref: v3.9.10 + - uses: actions/setup-python@v5 + with: + python-version: '3.11' + - name: Install CPU test dependencies + run: | + python -m pip install --index-url https://download.pytorch.org/whl/cpu torch + python -m pip install -r requirements.txt numpy safetensors pytest psutil opencv-python-headless Pillow aiohttp + - name: Compile tag + run: python -m compileall -q . + - name: Linux full regression on published tag + if: runner.os == 'Linux' + run: python -m pytest -q -p no:cacheprovider + - name: Windows storage regression on published tag + if: runner.os == 'Windows' + run: python -m pytest -q -p no:cacheprovider tests/test_temporary_storage.py tests/test_storage_api.py tests/test_p20_file_backed_tensor_poc.py tests/test_p22_file_backed_buffer.py tests/decode_cache + publish: + needs: [tag, tag-tests] + if: needs.tag.outputs.publish == 'true' + runs-on: ubuntu-latest + permissions: + contents: write + steps: + - uses: actions/checkout@v4 + with: + ref: v3.9.10 + persist-credentials: false + - uses: actions/setup-python@v5 + with: + python-version: '3.11' + - name: Build GitHub ZIP and verify source manifests + run: | + python - <<'PY' + import hashlib,pathlib,subprocess,zipfile,tomllib + root=pathlib.Path('.') + assert tomllib.loads((root/'pyproject.toml').read_text())['project']['version']=='3.9.10' + for name in ['MANIFEST.sha256','REGISTRY_MANIFEST.sha256']: + rows=(root/name).read_text().splitlines() + for row in rows: + h,p=row.split(' ',1);assert hashlib.sha256((root/p).read_bytes()).hexdigest()==h,p + print(name,len(rows),'matched') + files=subprocess.check_output(['git','ls-tree','-r','--name-only','HEAD'],text=True).splitlines() + target=root/'ComfyUI-H3-Continuum-v3.9.10.zip' + with zipfile.ZipFile(target,'w',zipfile.ZIP_DEFLATED,compresslevel=9) as z: + for p in files:z.write(root/p,'ComfyUI-H3-Continuum/'+p) + with zipfile.ZipFile(target) as z: + assert z.testzip() is None + assert set(z.namelist())=={'ComfyUI-H3-Continuum/'+p for p in files} + assert all(z.read('ComfyUI-H3-Continuum/'+p)==(root/p).read_bytes() for p in files) + pathlib.Path('SHA256SUMS.txt').write_text(hashlib.sha256(target.read_bytes()).hexdigest()+' '+target.name+'\n') + print('GitHub ZIP exact:',len(files),'files') + PY + - name: Install pinned Registry CLI + run: python -m pip install comfy-cli==1.22.0 + - name: Build and verify Registry archive + run: | + comfy --skip-prompt --no-enable-telemetry env + comfy node pack + python - <<'PY' + import hashlib,pathlib,zipfile + root=pathlib.Path('.') + declared={p:h for h,p in (r.split(' ',1) for r in (root/'REGISTRY_MANIFEST.sha256').read_text().splitlines())} + with zipfile.ZipFile('node.zip') as z: + assert z.testzip() is None + assert {i.filename for i in z.infolist() if not i.is_dir()}==set(declared)|{'REGISTRY_MANIFEST.sha256'} + assert all(hashlib.sha256(z.read(p)).hexdigest()==h for p,h in declared.items()) + assert z.read('REGISTRY_MANIFEST.sha256')==(root/'REGISTRY_MANIFEST.sha256').read_bytes() + print('Registry ZIP exact:',len(declared),'payload hashes matched') + PY + - name: Publish GitHub Release + env: + GH_TOKEN: ${{ github.token }} + run: gh release create v3.9.10 ComfyUI-H3-Continuum-v3.9.10.zip SHA256SUMS.txt --verify-tag --title 'v3.9.10 — Spectrum reuse and Reference assignment display' --notes-file docs/RELEASE_NOTES_V3910.md --latest + - name: Download and verify public GitHub assets + env: + GH_TOKEN: ${{ github.token }} + run: | + gh release download v3.9.10 --dir public-download --pattern 'ComfyUI-H3-Continuum-v3.9.10.zip' --pattern SHA256SUMS.txt + python - <<'PY' + import pathlib,hashlib + for name in ['ComfyUI-H3-Continuum-v3.9.10.zip','SHA256SUMS.txt']: + assert pathlib.Path(name).read_bytes()==(pathlib.Path('public-download')/name).read_bytes(),name + print('Public GitHub assets match verified local ZIP and checksum') + PY + - name: Publish verified Registry package + env: + REGISTRY_ACCESS_TOKEN: ${{ secrets.REGISTRY_ACCESS_TOKEN }} + run: | + test -n "$REGISTRY_ACCESS_TOKEN" + comfy node publish --token "$REGISTRY_ACCESS_TOKEN" + - uses: actions/upload-artifact@v4 + if: always() + with: + name: publication-v3.9.10 + path: | + ComfyUI-H3-Continuum-v3.9.10.zip + SHA256SUMS.txt + node.zip diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml new file mode 100644 index 0000000..7abbf11 --- /dev/null +++ b/.github/workflows/tests.yml @@ -0,0 +1,42 @@ +name: tests + +on: + push: + pull_request: + +permissions: + contents: read + +jobs: + test: + runs-on: ubuntu-latest + steps: + - uses: actions/checkout@v4 + - uses: actions/setup-python@v5 + with: + python-version: "3.11" + cache: pip + - name: Install test dependencies + run: | + python -m pip install --upgrade pip + python -m pip install --index-url https://download.pytorch.org/whl/cpu torch + python -m pip install -r requirements.txt numpy safetensors pytest psutil opencv-python-headless Pillow aiohttp + - name: Compile Python + run: python -m compileall -q . + - name: Run tests + run: python -m pytest -q -p no:cacheprovider + + windows-storage: + runs-on: windows-latest + steps: + - uses: actions/checkout@v4 + - uses: actions/setup-python@v5 + with: + python-version: "3.11" + cache: pip + - name: Install CPU storage test dependencies + run: | + python -m pip install --index-url https://download.pytorch.org/whl/cpu torch + python -m pip install numpy safetensors pytest psutil aiohttp + - name: Windows file-backed and temporary storage regressions + run: python -m pytest -q -p no:cacheprovider tests/test_temporary_storage.py tests/test_storage_api.py tests/test_p20_file_backed_tensor_poc.py tests/test_p22_file_backed_buffer.py tests/decode_cache diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..4e4fe86 --- /dev/null +++ b/.gitignore @@ -0,0 +1,6 @@ +__pycache__/ +.pytest_cache/ +*.py[cod] +*.safetensors +!tests/fixtures/v39_sampling_v5/**/*.safetensors +*.tmp diff --git a/AGENTS.md b/AGENTS.md new file mode 100644 index 0000000..55d3ffc --- /dev/null +++ b/AGENTS.md @@ -0,0 +1,42 @@ +# Contributor rules + +## Scope + +MiniMax H3 Continuum for ComfyUI. Package 3.9.10 provides the V3.9 Sampler and Reference Images helper, while retaining the separate V3.8X2 workflow. Source publication does not imply a GitHub Release or Registry package publication. + +Sampling Contract v6 and Graph Contract v4 apply to new generation. Keep original v5 history readable, without reusing its Takes in v6 generation or rewriting stored manifests. Public node IDs, socket/widget order, official workflows, Run Storage v3 and State/Session formats remain unchanged. + +## Product and data contracts + +- Keep normal generation, Review, retry, continuation and finalization discoverable. Advanced controls must not rewrite hidden values or generation identity. +- Follow ComfyUI Core validation. Do not reject prompts, unknown models, wrappers or merged models solely because they are unknown. Unrecognized prompt syntax falls back to Fixed and passes through as entered. +- Keep legacy strict_compatibility inputs loadable but ignore their values. Warnings remain diagnostic unless an unusable internal payload or corrupted data makes execution impossible. +- Preserve accepted chunks and prior Takes. State and Session remain safetensors plus JSON, written atomically. +- Keep full accepted chunk latents on CPU. Do not accumulate generated chunks in VRAM or decode Video/Audio VAE between chunks. +- Preserve native PackedLayout behavior and position_ids identity. Never silently resize continuation State or Session. +- Keep Reference conditioning, First/Last Image, Driving Audio and generated Audio contracts separate. Second Pass must preserve First Pass audio when the selected path requires passthrough. +- Keep SageAttention, Sol-Attn and Spectrum external and use their public wrappers. Do not replace Core classes globally or import private Spectrum runtime functions. +- UI nodes delegate to reusable modules. Preserve legacy implementation modules while current registered nodes import or inherit them. +- Main UI disclosure is presentation-only. It must not change backend kwargs, Sampling Contract or Run Storage identity. +- Keep standard Core display titles in public examples so users can identify node providers. +- Experimental changes default OFF and must not change Production behavior. Still Image Guide and deferred continuation interventions remain outside accepted Production scope. +- Keep V3.8X2 and V3.9 as distinct workflows. Do not automatically migrate historical node IDs, workflows, Runs or Takes. +- Keep the shipped workflow JSON/ZIP files byte-for-byte unless a workflow change is explicitly requested. + +## Validation and contribution + +- Preserve unrelated work and take a verified snapshot before changes. The snapshot destination is configurable through tools/snapshot.ps1. +- Reuse the existing pytest configuration, tests and runtime verifier. Distinguish CPU checks, offline registration, live loading, browser acceptance and GPU results. +- Runtime checks record prompts, media, model/LoRA/sampler settings, dimensions, chunk timings and observed results. Check saved workflow metadata when available. +- Verify all hashes declared by MANIFEST.sha256 and REGISTRY_MANIFEST.sha256 after modifying their payloads. +- Current public registration contains twelve nodes. Do not register experimental implementations implicitly. +- Do not claim whole-release acceptance from a targeted repair test. Browser save/reload and the unchanged default workflow remain separate acceptance checks. +- Keep machine-specific paths, credentials, service addresses, process identifiers and private operational records out of public documentation. +- Record public behavior and version history in CHANGELOG.md and WORKLOG.md. PROJECT_STATE.md describes the public source status. +- Commit and publish only when explicitly requested. Release/tag/Registry publication is a separate action. + +Run the standard validation entry point from the repository root: + +```powershell +powershell -ExecutionPolicy Bypass -File tools/validate.ps1 +``` diff --git a/CHANGELOG.md b/CHANGELOG.md new file mode 100644 index 0000000..4966baa --- /dev/null +++ b/CHANGELOG.md @@ -0,0 +1,374 @@ +# Changelog + +## 3.9.10 — Spectrum reuse and Reference assignment display (2026-10-20) + +- Normalize CRLF source line endings in both Spectrum file and module-closure audits. Keep the pinned source identities and existing Python 3.13 profile bytes unchanged. +- Add separate audited Python 3.12 profiles for the supported local and upstream 0.2.29 implementations. Keep unknown source/runtime state non-reusable, preserve old unsafe/v5 Takes, and require a new Run for newly supported configurations. Sampling v6, Graph v4, public nodes and workflow bytes are unchanged. No new GPU acceptance is claimed for this repair. + +## Unreleased — external Chunks Reference assignment table + +- Treat externally connected Chunks as unresolved in the V3.9 Reference Images + helper. Display assignment range 1–16 with the reason and explicit All/Off + range labels; retain direct-count tables and saved future assignments. +- Keep existing selectors, backend routing, Sampling/Run Storage contracts, + public inputs and official workflows unchanged. Loading or refreshing does + not automatically expand old selections. + +## 3.9.9 — upstream KJ Preview Override / Review continuation (2026-10-20) + +Fix new-Run v6 read-only with the separately audited upstream KJ profile, and +show saved reasons when reuse is unavailable. Keep the original local profile, +Sampling v6 guards, old Takes and official workflow bytes. + +CPU: **2,075 passed / 2 skipped / 2 known expected failures**, plus overlapping +90-case JS coverage. Upstream KJ GPU and normal-ComfyUI browser Continue both +passed: **Chunk 1 preserved, Chunk 2 generated, one reused / one new**; all four +AV outputs fully decoded. Browser checks used the existing local KJ profile. +These are targeted repair checks, not general quality or all-provider acceptance. + +Restart ComfyUI and hard-refresh after updating. On Preview Override, disconnect +optional VAE/Audio VAE and set Tiny VAE `none`; retain generation VAEs. Correct +unsupported settings before starting a new Run. Old unsafe/v5 Takes remain +readable and are not automatically made reusable. No workflow rewiring or Take +deletion is required. [Full release notes](docs/RELEASE_NOTES_V399.md). + +## Unreleased — V3.9 Turbo Sample (2026-10-16) + +- Add a separate user-supplied Auto / 2 x 8s Turbo dance and Thai rap sample, gallery copy, ZIP and setup notes. Preserve standard workflows, saved Takes and product code. This source-only sample has no new GPU acceptance and does not replace v3.9.8 release assets. + + +## 3.9.8 — Issue #31 Preview Override Phase A (2026-10-16) + +- Post-release documentation: after installing and restarting published tag `58fbd855`, both PyTorch/KJ and Kitchen + audited local Spectrum/KJ passed new-Run, separate-Queue Chunk 1 to 2 continuation with retained previews, unchanged prefix/contracts and four fully decoded AV outputs. Record the limited audited profile and unverified browser/quality/performance scope. Keep the original tag/assets and existing Takes/workflows. + +- Add a finite KJ Model Preview Override signature for the explicitly audited source profile. Preserve fixed preview settings, loaded implementation identity and cross-key/list wrapper execution order; omit only the reviewed builtin `frames` list. +- Limit reuse support to VAE/Audio VAE disconnected, KJ Tiny VAE `none`, and Core NoPreviews/Auto/Latent2RGB. Unknown source, altered helpers, additional state and approximate-VAE preview remain non-reusable. Generation itself retains the existing fallback. +- Keep unaffected MODEL signatures, Sampling v6 guards, old unsafe/v5 history and official Workflow bytes unchanged. Full real-Core CPU gate: 1,995 passed / 3 skipped / 2 existing Inspector expected failures / zero unexpected failures and CUDA initialization attempts; the 48 dedicated KJ tests passed without skips. Normal ComfyUI GPU acceptance passed four Core-API Queues: PyTorch and Comfy Kitchen INT8 + the audited local Spectrum implementation, each with separate-Queue Chunk 1 to 2 continuation and held preview images. Both second Queues reused one chunk and generated one; original Chunk 1 bytes and contracts were unchanged. All four video/audio outputs fully decoded. See docs/KJ_PREVIEW_OVERRIDE_V398.md for exact support and limits. + +## 3.9.7 — external-input regeneration cache (2026-10-15) + +- Fix automatic Full Run variations when Chunks or Variation Nonce come from external nodes. Unresolved linked inputs no longer let Core skip an explicitly requested regeneration. Earlier accepted chunks, existing Takes, public inputs and official workflows are preserved. Update the custom nodes, restart ComfyUI and refresh the browser; no workflow changes or history deletion are required. +- CPU: 1,948 passed / 2 skipped / 2 known Inspector expected failures; zero unexpected failures and zero CUDA initialization attempts. Targeted Core-API GPU queues verified external Chunks=2 / Nonce=0, repeated identical automatic variations 1 then 2, and 1 reused / 1 generated chunk per variation. Browser save/reload and Run-button submissions were checked separately. GPU scope: PyTorch attention, Spectrum OFF, 512x512, 8-step Turbo, two 5-second chunks. This is a targeted cache repair, not general image-quality or all-accelerator acceptance. + +## 3.9.6 — regeneration intent and automatic variation cache (2026-10-15) + +- Preserve explicit Regenerate From and Variation Nonce across V3.8/V3.9 workflow restoration and Queue submission. Persist only the selected intent as workflow metadata; keep the legacy unowned Chunk 1 / nonce 0 cleanup and one-shot restart behavior. +- Execute Full Run explicit regeneration with automatic nonce 0 through Run Storage on every Queue. Ordinary Full Run and fixed positive nonces retain Core caching. Sampling v6 safety, accepted Takes, public inputs and official workflow files are unchanged. +- Full CPU gate: 1,908 passed / 2 skipped / 2 known Inspector expected failures; zero unexpected failures and CUDA initialization attempts. Normal ComfyUI browser save/reload retained Chunk 2 / nonce 5 and Chunk 1 / nonce 0 after history loading. Five new-Run GPU queues passed continuation, Chunk 2 regeneration and two identical Full Run / nonce 0 submissions; earlier Chunk 1 bytes were preserved. GPU scope: PyTorch, Spectrum OFF, 512x512, 8-step Turbo; all five media outputs fully decoded. See docs/REGENERATE_INTENT_V396.md. + +## 3.9.5 — Issue #31 Attention and Spectrum reuse signatures (2026-10-15) + +- Observe the loaded Core PyTorch and Comfy Kitchen INT8 attention overrides with a finite local descriptor, including the verified wrapper self-reference, implementations, defaults, closure, both container dispatch paths, numerical helpers and provider state. +- Keep Sampling v6, runtime signature v1 and Graph v4 unchanged. Overrides outside this adapter retain their existing observation policy; unrelated incomplete MODEL/CFG observations still disable reuse. +- Add a dedicated Spectrum binding identity for explicit audited provider implementations, including upstream Spectrum 0.2.29. Include every effective immutable config value and verified loaded functions/wrappers; exclude reviewed execution-local history and counters. Unknown provider sources, extra runtime state or altered callables remain non-reusable. +- Preserve original unsafe Takes and v5 history. Only new correctly identified runs can gain reusable prefixes. Public nodes, workflow bytes and storage schemas are unchanged. See the versioned validation record for CPU/GPU coverage and remaining limits. + +## 3.9.4 — Registry packaging repair (2026-10-14) + +Include the declared Review restoration guide and exclude repository-only prompt-authoring files from the Registry ZIP. The publishing workflow now builds the actual package and checks its exact file set, CRC and every Registry Manifest hash before uploading. Runtime behavior, twelve public node IDs, workflows and saved Runs/Takes are unchanged. GitHub Releases 3.9.3 and 3.9.2 remain available. + +## 3.9.3 — temporary storage settings (2026-10-13) + +- Add a server-owned scratch folder setting for both Finalize and Decode Cache, with storage candidates, logical file sizes, free space and conservative unused-file cleanup. +- Apply changes to new files while keeping existing outputs and aliases valid. Preserve blank/default paths and explicit internal overrides; never fall back silently to the original disk when a selected drive is unavailable. +- Protect unknown owners, live processes and mapped outputs during cleanup, including outputs whose helper has already been removed. Track new cache directories for dead-process recovery. +- Add manual RAM and Disk-backed capacity checks and account for discoverable Linux cgroup memory limits. Keep Auto's 4 GiB threshold and reserve, existing node controls, frame dtype and generation/storage identities. +- Keep V3.9 and V3.8X2 workflows and public node schemas unchanged. No new GPU generation is claimed for this storage update; see PACKAGE_VALIDATION.txt for CPU/browser/platform coverage. +- Follow-up: report missing/empty/invalid storage responses with HTTP status and recovery guidance. Keep storage candidates visible at a usable width and mutations disabled until status is available. Preserve Core address/profile handling and never replay a mutating request. + +## 3.9.2 — Issue #30 frontend hotfix (2026-10-13) + +- Skip upstream Review settings snapshots when no comparison baseline exists. Keep drawing-time edit detection and synchronous Queue/Take checks for reviewed settings. +- Notify connected Samplers once per Reference selection through the native callback, and skip `setSize()` when node dimensions are unchanged. Native assignment values still update synchronously; saved workflow and generation contracts are unchanged. +- Track the separate Reference Plan Inspector refresh failure after repeated Sampler configuration as an expected regression failure. No delayed update timer or new animation-frame scheduling is introduced. +- Bump package metadata to 3.9.2 while keeping V3.9 node IDs and existing generation/storage contracts. Native values and Review safety guards remain synchronous. Mobile freeze resolution is unconfirmed; Reference table rebuilding and snapshots with a Review baseline remain. + +## 3.9.1 — stability fixes (2026-10-04) + +- Preserve audio-only and mixed Second Pass keyframes while spatially adapting only video. +- Use Sampling Contract v6 / Graph Contract v4 for safe Take reuse, including sampler closures and MODEL CFG/wrappers. Keep v5 history readable without adopting its prefixes into v6 generation or rewriting old manifests. +- Slice Review Driving Audio once from physical-group natural time; use Core resampling and document explicit Reference Encode Cache invalidation for direct VAE weight mutation. +- Match V3.9 partial Review conditioning and Reference routes to actual output-group identities before Second Pass. Preserve complete-run behavior, original audio LATENT objects and saved Takes. +- The repaired runtime passed 1696 CPU cases (1 skipped) and three targeted 512×512 Review Second Pass GPU cases. Browser acceptance remains pending. Optional latent-upscale doubled outlines are tracked separately in Issue #27 and are not a required gate for this stabilization. +- Bump package metadata to 3.9.1 while preserving V3.9 public node IDs, widget order and official workflows. Restart ComfyUI and refresh the browser after updating. Keep historical V3.8 releases untouched; this patch adds no LoRA Plan. Versioned source downloads are provided through GitHub Releases; Registry publication is separate. + +## 3.9.0 — V3.9 source on main (2026-09-29) + +- Publish the separate V3.9 Sampler and nine-slot Reference Images V3.9 helper with fixed `@R1`–`@R9` identities and All / Per chunk assignments. V3.8X2 keeps its existing Sampler and Reference connections; saved Runs/Takes are not automatically migrated. +- Ship the user-selected V3.9 workflow JSON unchanged, a V3.9-only workflow ZIP, and a combined V3.8X2/V3.9 workflow ZIP. The ComfyUI Templates gallery exposes both separate workflows. +- Promote package metadata and startup version to `3.9.0`. Keep historical Releases/tags untouched. This main-source commit is not a new GitHub Release or Registry publication; browser save/reload and the unchanged default template's GPU acceptance remain separate gates. + +## V3.9 development history — Reference Images in-node UI (2026-09-27) + +- Add a separate nine-slot Reference Images V3.9 helper with fixed `@R1`–`@R9` IDs, in-node All chunks / Per chunk table, and one typed connection to the V3.9 Sampler. Both choices use the existing group-effective routing engine; disconnected or unassigned slots do not renumber later images. +- Supply a separate V3.9 workflow JSON/ZIP and user guide. V3.8 Sampler and V38X2 workflow JSON/ZIP remain unchanged. Older V3.9 prototype graphs require manual rewiring and saved Runs/Takes do not migrate automatically. +- Local CPU `1530 passed / 1 skipped / 0 failed`, frontend Reference Images and Review Queue contracts, and standalone runtime verification passed. Live-browser and GPU acceptance of this new helper/UI remain pending. No commit, push, Release or Registry publication. +- Rebase the separate V3.9 JSON/ZIP on the supplied V38X2+Decode Cache Helper workflow, preserving its appended Repeat Reference video setting, media loader values, Image bypass and Decode Cache Helper. Route Reference Images 1–9 through the new helper; keep all other graph nodes and settings unchanged. Restart ComfyUI fully after updating to load the new node schema. + +## Prior local prototype — V3.9 Reference Routing RR-R6B (2026-09-25) + +- Add a separate V3.9 Sampler ID with default All (unchanged V3.8 Reference path) and opt-in Custom R1–R9 per-chunk selectors. Keep the V3.8 schema/widget order and official V3.8X2 workflow files unchanged. +- Carry the frozen RR-R5 group contract to a read-only, post-Queue Plan Inspector. Pending groups remain configured/unverified; accepted groups display backend-verified generated/reused status, effective Picture mapping, SHA and warnings. +- Capture matching group-local routed conditioning in same-Queue Refine Context; Second Pass distinguishes verified Reference inheritance from prompt-only fallback. No cross-session reconstruction is claimed. +- Local full CPU: `1524 passed / 1 skipped / 0 failed`. RR-R6 GPU/API integration passed 11 Queues, including Review Q1–Q3, Second Pass, and the Sage/Sol/Spectrum wrapper matrix. The subsequent real-browser B0–B5 Review, completion, and save/reload gate passed using a corrected Ref2V 8-step LoRA / eight-step workflow; this is not a six-step rerun. Subjective identity and long-form image quality remain ungraded. +- V3.8X2 workflow JSON/ZIP and the V3.8 Sampler remain available. V3.9 `All` uses the V3.8 shared-reference path; changing to V3.9 `Custom` is a manual workflow change and does not automatically migrate an accepted Run or Take. +- The planned separate V3.8X2 tag/Release and `release/3.8x2` maintenance branch have not been created. V3.9 workflow, commit, push, Release, and Registry publication remain pending. + +## Experimental on main — Unified Timeline Video (2026-09-22) + +- Opt-in Follow Timeline / legacy-compatible Repeat Reference on the existing IMAGE input. +- Per-physical-group reference selection; source-end fallback; distinct interval VAE keys and local prompt caches. +- Local acceptance passed: full CPU `1424 passed / 1 skipped / 0 failed`, browser save/reload, and Follow/Repeat `2 × 5 s` GPU runs at 704×416/24fps with synchronized 10-second audio/video, mode separation, and no OOM/NaN/crash. This remains Experimental and is not a frame-perfect motion-copying or long-video RAM-optimization claim. + +## Unreleased — Loader persistence repair (2026-09-21) + +- Repair Issue #23 by delegating inherited native mode accessors; preserve Enable Image and remove draw/save-time widget collection replacement. +- Switch the two paired official V3.8X2 workflows to Core Load Audio and Core Load Video plus an Enable-free H3 Continuum Video Adapter using the existing FPS converter. +- Keep old Audio/Video IDs as deprecated compatibility nodes. Preserve other workflow nodes, settings and output contracts; regenerate matching ZIPs and integrity hashes. +- Add store-backed loader and converter regressions. Windows/browser acceptance remains a separate pending check; no Release/tag/Registry publication. + + +## 3.8.3 — V3.8X2 release preparation (2026-09-19) + +- Vendor the previously standalone Decode Cache Helper implementation into Continuum as a separate `H3DecodeCacheHelper` public node. Preserve its node ID, inputs, outputs, list behavior, native Core Decode delegation, and process-local cache behavior. +- Add `H3 Continuum Reference Images` for optional Reference Images 4–9 while retaining the Sampler's direct Reference Images 1–3 and existing 0–3 image workflows. +- Keep Sampler, Finalize, Assembly Plan, Core Decode and existing V3.8X widget/socket contracts unchanged. +- Add the accepted workflow under two official distribution names: `MiniMax_H3_Continuum_V38X2` and `MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper`. The files contain the same graph and are not separate workflow variants. +- Document the nine-reference GPU/RAM gate and recommend pre-scaling large reference sources to about 0.30 MP when using all nine on a 16 GB GPU. +- Clarify that Decode Cache Helper speeds only repeated Decode of unchanged latents. It does not speed Sampling, a first MISS may be slower, and the measured 20%+ reductions are configuration-specific rather than a general guarantee. + +## 3.8.1 — V3.8X release candidate (2026-09-15) + +- Promote the accepted R0–R6 Runtime Redesign and final local GPU Functional Gate to the V3.8X (`3.8.1`) release candidate. +- Rename the current supplied Spectrum workflow to `MiniMax_H3_Continuum_V38x.json` and its matching ZIP. The JSON and ZIP remain byte-for-byte paired; the graph is unchanged. +- Preserve old V3.8.0 installation through the existing `v3.8.0` Git tag. Do not mix its workflow with V3.8X; use the matching tag/package for an older saved workflow. +- No new public node, socket, widget-order, Sampling, Review, or Run Storage contract is introduced by this release label. +- Compatibility note: validated V3.8X with external ComfyUI-Spectrum-MiniMax-H3 v0.2.27 in a Sage + Spectrum 3 x 5-second run. It completed without the previous `aimdo memory compile error`; decoded video/audio integrity and chunk-boundary checks passed. This relies on Spectrum's own v0.2.25+ AimDo malloc-graph fix; Continuum does not vendor or modify Spectrum. + +## Unreleased — V3.8.x release hygiene (2026-09-14) + +- Updated the canonical V3.8 workflow to remove both `easy cleanGpuUsed` nodes. First Image now connects directly to the Sampler, and Core Create Video connects directly to Core Save Video; ComfyUI-Easy-Use is no longer a workflow dependency. +- Fixed the shipped legacy `v3.reference_video` module to import the shared temporal helper from the package root. +- Hardened the Windows installer: validate and copy only the Registry Manifest payload, and move previous installations to a backup directory outside `custom_nodes`. +- No Sampling, Review, or Run Storage behavior is changed by these release-hygiene repairs. + +## 3.8.0 — Review UI hotfix on main (2026-09-08) + +- Withdraw stale Continue/Retry/Finish actions after local or connected generation inputs change. A stale queued Retry falls back to normal backend compatibility/reuse checks instead of targeting an incompatible saved review. +- Add the frontend-only `Start again from Chunk 1` action: use existing non-destructive branch controls, preserve saved Takes, and consume the restart selection once after Queue. +- Preserve explicit Regenerate From boundaries instead of resetting them when a Review action is clicked. Preserve normal continuation, completed review, history and serialized widget order. +- No sampling, audio, continuation transport, workflow or public Python schema change. The separate 15-second Review static/noise report remains unresolved. This main-branch hotfix is not a new Registry publication or release asset. + +## 3.8.0 + +### Added + +- Established the seven-node V3.8 public surface: Sampler V3.8, Finalize, Load Image, Load Audio, Load Video, Second Pass, and the modular Reference Audios helper. Unreleased Easy facade prototypes remain in source but are not registered. +- Added `H3 Continuum Sampler V3.8` with shared `Draft` 0.30 MP, `Balanced` 0.60 MP, `Native 768`, and `Custom MP` resolution presets. +- Added `Review Each Chunk`, `Continue / Next`, `Regenerate Current`, and `Finish Remaining` controls backed by Run Storage review branches and canonical-head recovery. +- Publish one user-supplied V3.8 Spectrum workflow as matching JSON and ZIP. It requires Spectrum, rgthree, and KJNodes; the same graph can be switched to a task-matched LightX2V Turbo LoRA. +- Added Continuum Image, Audio, and Video loaders with explicit native bypass behavior for optional workflow inputs. +- Added `H3 Continuum Reference Audios`, which bundles up to three ordered Core-native standalone audio references behind one appended V3.8 Sampler socket while preserving the legacy single-reference path. +- Replaced the Main `Landscape / Portrait / Square` choice with a two-path `Size Source`: preserve the connected First Image aspect at the selected preset, or enter an exact aligned Manual Width and Height. + +### Improved + +- Integrated Driving Audio with Review, Smart Regenerate, Finish Remaining, and restart persistence. +- Preserved FL2VA Long Terminal Merge as one atomic physical review unit, including Smart Regenerate of the complete terminal pair. +- Reused one resolution resolver across Easy and normal V3.8 samplers while leaving the V3.7 schema and Production path unchanged. +- Extended Run Storage with Review lifecycle metadata, prefix reuse, nonce branches, and latest canonical-head restoration without changing the sampling identity contract. +- Added a frontend-only `Basic / Production` node property for Sampler V3.8. It changes presentation only and preserves the Python schema, widget order and values, backend inputs, Sampling identity, and Run Storage identity. +- Added Registry packaging rules that exclude tests, tools, diagnostics, Labs assets, and historical workflows while retaining all runtime-imported implementation modules. +- Kept the old V3.8 Aspect input as a hidden compatibility field. Saved Auto workflows migrate to First Image, while saved Landscape, Portrait, and Square workflows migrate to the same resolved Manual canvas. + +### Notes + +- Second Pass / `refine_context` is not supported on a partial Review sequence. Complete Review before running Second Pass. +- The 16GB GPU Gate passed with small headroom: Strict Minimal peaked at `14.96 GiB`, and complex accepted gates reached about `15.5-15.6 GiB`. Results vary by environment and are not a universal 16GB guarantee. +- Still Image Guide remains Experimental. V3.7 and earlier Node IDs are not exported by V3.8; workflows using them can appear with unknown nodes and should be opened with the matching historical Release/tag. + +## 3.7.0 + +- Added `H3 Continuum Sampler V3.7` with one optional Still Image Guide input while preserving the V3.6 required-input schema, Production defaults, and saved-workflow compatibility. +- Generalized the Second Pass Conditioning Adapter so First/Last images, continuation context, and an owned Still Image Guide can be rebuilt at target latent geometry. Physical-group and Long Terminal Merge ownership remain intact. +- Added the internal schema-v1 `RefineSchedule` contract with `External`, `Full`, `Tail`, and `Partial` modes. Existing SIGMAS input remains compatible, and Tail/Partial use exact ranges from the source schedule. +- GPU-validated the internal Tail 6 suffix at 704×704 against the prior external Tail 6 path, with decoded Video SSIM `0.982936` and bit-exact decoded Audio PCM. +- Completed a 1152×1152 Conditioning correctness gate in which the original Still Image Guide source was reencoded at target geometry without changing the first-pass Audio contract. +- Added explicit source-role ownership so a Still Guide sharing the same frame as First/Last is not replaced by the endpoint image, and restricted Still Image Guide input to exactly one IMAGE. +- Added V3.7 compact-frontend parity for project identity, hidden internal widgets, Run Storage controls, queue normalization, diagnostics, and permanent Reference Audio sockets. +- Kept Still Image Guide **Experimental / Production HOLD**. Core hard-anchor semantics can cause an abrupt trajectory change at the Guide frame and redirect later motion; it is not a smooth transition control. +- Preserved V3.6/V3.5/V3.4 node IDs, public Second Pass schemas, Core PackedLayout/RoPE construction, Run Storage compatibility, and Production generation defaults. + +## 3.6.1 + +- Canonicalized an absent Last Frame hash as `""` and added a narrowly scoped compatibility path for V3.6.1 caches whose old final chunk stored `"none"`. Extending a normal no-Last-Frame run from two to three chunks now reuses both stored chunks, while a real Last Frame and Long Terminal Merge still invalidate/regenerate their required sampling units. +- When Run Storage is turned Off, the compact frontend now resets hidden `Regenerate From` to `Auto` and `Variation Nonce` to `0`; the backend rejection remains as a non-UI safety check. +- Added non-blocking mixed Prompt syntax diagnostics. Timeline mode now ignores standalone `---` separator lines with `H3C-P105`, while preserving existing uncovered-chunk fallback and Fixed-prompt fail-open behavior. +- Prevented V3.6 Standard continuation from stopping when Audio Continuity is enabled with Fast 5, Strong 39, or Auto. These modes now resolve to the accepted Reference Context transport before Run Storage identity is created; Balanced 22 continues to use Masked AV. +- Kept Standard with Audio Continuity disabled on Masked Video, and kept Compatibility on Reference Context for every Continuity mode. +- Added explicit resolved-transport diagnostics without rewriting the user's Backend or Continuity selections. +- Preserved V3.5.3/V3.4 nodes, Sampling, Terminal Merge, Assembly, Seam, Run Storage schema, and saved-workflow compatibility. + +## 3.6.0 + +- Added `H3 Continuum Sampler V3.6` with Standard Masked AV continuation. Prior finalized Video/Audio prefixes are preserved inside the next H3 target with Core noise masks instead of being appended as duplicate Reference rows. +- Retained `Compatibility` as an Advanced fallback to the V3.5 `Reference Context` transport. Internal transport identifiers are not exposed in the public UI, and all V3.4/V3.5 node IDs and saved workflows remain loadable. +- Scoped Run Storage identity and Sampling Contract execution semantics by nondefault continuation backend, preventing Reference and Masked chunks from being mixed while preserving legacy Reference revision identity. +- Preserved Long Terminal Merge as one physical sample during resume and `Regenerate From`; logical chunks `[2,3]` are never split or partially reused. +- Extended the shared V2/V3 `chunk_seconds` range to 4–30 seconds with a 5-second default and 0.1-second step. The 5–15 second range remains recommended and validated; longer high-resolution chunks can substantially increase VRAM and runtime. +- GPU-accepted T2VA, I2VA, Reference Image + Reference Audio, Balanced 22f/37T Joint AV, and 3×5-second FL2VA Long Terminal Merge Masked continuation. Protected Video/Audio prefixes finalize bit-exact, and the accepted PIG-3 audio output passed subjective listening. +- Measured the accepted FL2VA Terminal Group 2 path at 8.01% fewer packed rows and a 4.06% lower median Sampling time than Reference Context in the tested paired runs. These are environment-specific measurements, not universal speed guarantees. +- Preserved V3.5 Second Pass, Hi-Res Fix, Conditioning Bridge, disk-backed Assembly, Prompt/CLIP cache, Video Guide optimization, V3.4 compatibility, and existing Audio/Video Seam algorithms. + +## 3.5.3 + +- Repaired the public Conditioning Bridge workflow's Width/Height input slots and added serialization-integrity checks for every published workflow. +- Removed stale orphan link IDs from the public V3.4, V3.4 Turbo, and V3.5 workflow files. +- Corrected Hybrid First/Last + Reference warning messages so public `` numbering remains absolute and matches the accepted runtime behavior. +- Losslessly re-encoded one unreferenced legacy PNG that strict image decoders could not read. +- Preserved all V3.5.2 generation, Sampling, Conditioning, Terminal Merge, Assembly, Seam, Run Storage, Prompt/CLIP cache, Video Guide optimization, and V3.4 compatibility contracts. + +## 3.5.2 + +- Added a conservative V3.5-only Prompt/CLIP cross-run LRU cache for unchanged T2VA, I2VA, and FL2VA conditioning. Reference, scheduled, tokenizer-option, and hook-modified CLIP paths bypass the cache; V3.4 is unchanged. +- Verified cache HIT execution with `encode_calls=0` and bit-identical decoded video and audio PCM. The optimization removes repeated Prompt/CLIP work only; Sampling still runs normally. +- Reduced long Video Guide preprocessing memory by hashing and finite-checking the complete source in eight-frame CPU chunks while retaining only the prefix required by H3 conditioning. +- Preserved the exact Video Guide source SHA-256, Run Storage identity, `17k+5` frame alignment, final-frame padding, VAE input, Qwen input, Sampling seed, decoded video, and decoded audio. +- Measured the deterministic CPU comparison at approximately 71% lower additional peak RSS (`396.8 MiB` to `114.7 MiB`) and 59% lower retained storage (`225 MiB` to `93 MiB`) with effectively unchanged preprocessing time. +- Completed the Phase 3/4 performance and feature-overhead audit. No speculative Sampling, Driving Audio lazy-decode, Audio hash, Assembly, Seam, Session, or V3.4 optimization was adopted. +- Preserved all V3.4/V3.5 node IDs, backend socket keys, saved-workflow compatibility, Sampling, Conditioning, Terminal Merge, Assembly, Seam, and Run Storage contracts. + +## 3.5.1 + +- Added `H3 Continuum Conditioning Bridge V3.5`, exposing one complete Core-compatible MODEL and CONDITIONING object per Continuum physical group without flattening CONDITIONING entries. +- Added optional standalone Reference Audio conditioning to Sampler V3.5. Its two Python-defined sockets remain permanently present; the frontend no longer adds/removes them or serializes a visibility-only widget, preserving workflow widget alignment on save/reload. +- Removed the experimental Last Queued Seed override and restored standard ComfyUI `control after generate` behavior for the V3.5 sampler. +- Renamed the displayed video-guide inputs to `Video Guide Frames` and `Video Guide Size` to distinguish a video loader's IMAGE frame batch from still Reference Images. Backend input keys and saved-workflow links are unchanged. +- GPU-accepted Conditioning Bridge generation, Reference Audio/Image retention, and the V3.5 Second Pass/Hi-Res paths while preserving physical-group, Terminal Merge, Run Storage, and audio contracts. + +## 3.5.0 + +- Added `H3 Continuum Sampler V3.5` with a Continuum-aware refine-context output while preserving the V3.4 sampler path. +- Added `H3 Continuum Second Pass V3.5` as the advanced bridge for externally processed H3 video latents, including physical-group and Long Terminal Merge preservation. +- Added experimental `H3 Continuum Hi-Res Fix V3.5`, using a bounded pixel/VAE resize round trip before low-denoise Second Pass sampling. +- Added `H3 Continuum Latent Resize V3.5` as an advanced utility; direct high-ratio latent interpolation is not the recommended main Hi-Res Fix path. +- Added `H3 Continuum Assemble + Seam V3.5` with Auto, RAM, and Windows-safe disk-backed video buffers while keeping audio in RAM. +- Added the recommended V3.5 template workflow with optional Hi-Res Fix, Auto assembly, shared still-image megapixel sizing, and separate Video Guide Frames sizing. +- Preserved Exact Duration, seam behavior, physical decode-group ordering, first-pass audio passthrough, and external Core VAE Decode contracts. +- Kept V3.4 nodes and saved-workflow compatibility unchanged. +- Validated a 9.49 GiB final IMAGE with hash-identical RAM/Disk-backed output and approximately 9.50 GiB lower Windows private memory in Disk-backed mode. +- GPU-accepted the integrated 1x5 576-to-1152 Main path and Hybrid/Reference 1x5 paths; documented that the 3x5 2x terminal 77T group exceeds the tested RTX 5060 Ti 16 GiB. + +## 3.4.0 + +- Completed the V3.4 runtime path from the public sampler through V3/V2 to `run_sequence()`. +- Added Driving Audio with absolute-time guide slices while preserving the original source for final output. +- Added persistent Video Guide Frames conditioning with Efficient 0.4 MP, Balanced 0.6 MP, and Match Output modes. +- Included Driving Audio and Video Guide Frames identities in Run Storage contracts. +- Changed malformed or empty prompts to warning-and-fallback behavior instead of stopping generation. +- Removed independent compatibility, model, and upstream-node restrictions that were stricter than ComfyUI Core. +- Added V3.4 runtime, driving-audio, and context diagnostics coverage. + +## 3.3.0 + +- Promoted Timeline Video to a stable public node with chunk-local slicing and a 0.4 MP default. +- Added `H3 Continuum Assemble + Seam` with stable Audio Seam Auto and guarded Video Seam Auto defaults. +- Kept `Auto 2` as an experimental exposure-ramp extension without widening the visible option label. +- Added native Reference Audio conditioning and retained up to three ordered Reference Images. +- Preserved stable sampler Node IDs, Run Storage compatibility, Core VAE decode, and Spectrum Interop API v1 Actual Prefix 2 behavior. +- Added consolidated runtime validation results and documented known Timeline Video and Auto 2 limitations. + +## 3.2.4 + +- Added native MiniMax H3 Reference Audio 1 conditioning with deferred Audio VAE encoding. +- Matched Core's permissive prompt handling by warning instead of stopping on unavailable Picture or Audio tags. +- Matched Core's dynamic Reference behavior by compacting active image inputs into Picture 1 through Picture N. +- Kept Sampling, Continuation, Spectrum Interop, external Core VAE Decode, and assembly semantics unchanged. + +## 3.2.3 + +- Formalized the third ordered Reference Image as a compatibility-preserving extension. +- Kept V3.2.2 zero, one, and two-image Reference Contract JSON unchanged. +- Added Image 3 shape, dtype, exact SHA-256, position, and preprocess metadata only for three-image runs. +- Added Golden Contract and Revision identity regression coverage for V3.2.2 saved runs. +- Kept Sampling, Continuation, Spectrum Interop, Core VAE Decode, and assembly semantics unchanged. + +## 3.2.2 + +- Allowed Ref2VA, FL2VA, and unverified checkpoints with Reference conditioning. +- Kept checkpoint classification as diagnostics without automatic MODEL switching. +- Kept Strict Compatibility for genuinely unsafe or unsupported H3 contracts. + +## 3.2.1 + +- Declared V3.2.1 stable after GitHub Actions, Windows Ref2VA generation, Spectrum Actual Prefix 2, chunk integrity and complete Auto Resume passed. +- Added T2VA, First/Last Frame, FL2VA and persistent two-image Reference conditioning to the production sampler. +- Added crash-safe raw AV Run Storage, automatic resume, deterministic Revisions and partial regeneration. +- Combined ordered upstream graph fingerprints with runtime MODEL/CLIP/VAE weight probes for fail-closed reuse. +- Added per-chunk SHA-256 verification and rejected Regenerate From when Run Storage is disabled. +- Added official MiniMax H3 Sigma Shift graph-contract support. +- Updated CI dependencies, Windows runtime verification and package metadata for V3.2.1. +- Standardized project licensing as MIT. +- Kept sampling, Continuation, Spectrum Interop and external Core VAE decode semantics unchanged. + +## 2.1.7 + +- Added the compact static `H3 Continuum Sampler` facade over the unchanged V2 execution core. +- Added Clip Overrides, Advanced, and Result pack helper nodes. +- Preserved `H3ContinuumSamplerV2` as a Legacy/Core workflow compatibility node. +- Removed the frontend display controller and all dynamic socket, proxy-widget, DOM, resize, and serialization workarounds. +- Kept State, Session, Prompt Plan, Sampling, Native Continuity, and Spectrum Interop contracts unchanged. + +## 2.1.6 + +- Rebuilt the Sampler V2 frontend as one Vue-aware display controller. +- Removed the Individual Clip Overrides accordion; only active Clip Prompt sockets are shown. +- Kept connected out-of-range Clip Prompt sockets visible when chunks is reduced. +- Added one Advanced Settings control; connected advanced sockets stay visible while collapsed. +- Removed fixed node-height handling and all input/output add/remove UI folding. +- Kept backend socket/widget order, types, serialization, State/Session/Prompt Plan schemas, sampling, and Spectrum Interop unchanged. +- Exposed show_preview as a per-node basic control and removed the duplicate global preview setting. + +## 2.0.1 + +- Fixed the bundled V2 example workflow widget serialization for `base_seed` with `control_after_generate`. +- Added an in-memory migration guard for the malformed 2.0.0 example where later widgets were shifted by one slot (`diagnostics` arrived as `0`). +- Preserved valid 2.0.0 workflows and all V1/V2 node identifiers and schemas. +- Added regression coverage for the legacy shifted-widget signature. + +## 2.0.0 + +- Added production **H3 Continuum Sampler V2** for integrated N-chunk generation. +- Kept the V1 continuation algorithm and all V1 node identifiers unchanged. +- Added Fixed, List, and Timeline prompt planning plus a connectable Prompt Plan node. +- Cached each unique Qwen prompt conditioning before H3 sampling. +- Reused one externally accelerator-patched MODEL, sampler, and sigma schedule across all chunks. +- Deferred Video/Audio VAE decoding until every H3 chunk finished. +- Captured full raw AV chunk latents and next-state tails on CPU before decoding. +- Removed the second device-to-host state transfer by deriving next-state tails from committed CPU entries. +- Added conservative latent-motion Auto Context selection with 5/22/39-frame profiles. +- Added deterministic independent 64-bit per-chunk seeds and reroll nonces. +- Added V2 Session resume, accepted-prefix reuse, branch-safe reroll, and atomic safetensors/JSON persistence. +- Added exact cumulative audio sample-boundary assembly to prevent long-chain rounding drift. +- Added exact total-duration correction. +- Added single-allocation decoded IMAGE assembly and preflight RAM safety checks. +- Added Basic and Full diagnostics, including video overlap MAE/PSNR and audio correlation/boundary jump. +- Added optional V1 State entry point and V2 Session-to-V1 State conversion. +- Added V2 public API modules and regression/orchestration tests. +- Updated installer runtime verification for all ten registered nodes. + +## 1.0.1 + +- Fixed a false strict-compatibility failure when Sol-Attn wraps MiniMax H3 + `PackedLayout.__init__` with `*args` / `**kwargs` for Morton/span tracking. +- Strict checking still fails closed when the native H3 constructor genuinely + removes required positional or keyword capabilities. +- Added regression tests for the native signature, Sol-Attn forwarding wrapper, + and a genuinely incompatible constructor. + +## 1.0.0 + +- Added **H3 Continuum Join** for initial and continuation clips. +- Added direct native video/audio latent state capture with 5/22/39-frame profiles. +- Represented continuation history as one native H3 `video` / `video_audio` reference block. +- Added Finish, Assemble, atomic State Save/Load, signed audio-grid offsets, in-place MM-RoPE adjustment, strict compatibility tests, and accelerator composition. diff --git a/H3-Continuum-LLM-System-Prompt-v1.zip b/H3-Continuum-LLM-System-Prompt-v1.zip new file mode 100644 index 0000000000000000000000000000000000000000..43ae8467c481607d291a36fb341e8f93896e2736 GIT binary patch literal 11595 zcmZ{KQ+Q@eyKQXeOFFjge6elYwr#WHbZna)I~{dw+jhsv>Af$`TKnw(%(|Gf>UnBh z)qKZ0Y79jga0mbh2nYPGkNJu)% zJUYTSm#lozbAClmoDm-`BA7Ss-S2gOyG8iax^u;fp+k}8m`c5bODCNvQc`E|YJJO9 z2e5(ZFNwmN6~QQ1I^`OsGzy{!@3RIeQjH696*`S^_t63qv0Ke$%Jb`kQNho;UK|uk@5vC}{8=o0JYEmQtD=eOHIt$YJ@T~=#>8X)a zEFZ6R+EyGqs}~?O;@A)AZ?y#85BD#-MHf3xT%?O|?xhpKZ zGr##eZqWV-P8k*%;E=$k{jjuqV48_L*MSCE$0clEq@d7-bfa%c1`2V=OCFq5wikPG z8H`cRr9z~ZLR*f}4rV2ItUTtJLEWlenG`rvm%6Yph>Zg+Nv>e{EC}v96koG|puH++ zNxm;GIqhcNxg})EKGlExy+Sq>2b(-eG1N_*d^2C_6fOhrbA;W=@y}C*duE&ERM@`w z&1iv|{uCLs9`)ZZkDYo?dc=!H&cnl(~b=MaB``N&n`g7kte)rma>_Ztp``#gX;*J86*oTA*vN9 z+?LCOhqs#dV^rP(&cCF#z`LmJpa_yuTYz?4KhKz7$Qf;o^{bx_xD?D6i9Q5eJs zhE*AE>MEfq;ZFPvIjI+*-I_r-PoO+`27Vn=G%lVyU@CGM&nt44R5VW6F=CWF6B|`E zcb2?PuT@)--v%k-?R@wymX=9L>+wAt&0FvJImI=RsE2fteEV^d z)^GoGk1YRbN8aagK=UV9#E$1BO~Il`trtkZm^Myyxh+C9t`xW3ivJNl((pQNBETgZ zKg}4KC-8n4c{?HfM7}+J={hv45>s;PzxC{p+g|BMe^=Sb#Jbv?h&U^=gp z!7hulrzpETNPE97D+rN(*G&Go_3gM&0QeKPK5`twPeWRc-|We?rYC33O0soO`YFG4 zf7Z;iOYb|GvZ=|=B)2asyA5zQ)`EehuQe_|i;%(9^7EFNhPQ)cgTz=PMQyCdh3jp2 zdjqm+QjC|{g7)U|jj-R>k|D$W$M705Q~`!o_~sXg;{UDwoqwzUM@%KfC=>`tJQ4^9 z;eV?BZg%wl#RJYpCa!;3pj%_o9+wNP=bRdD)CgU0^Tz=&V3+_vx%${A(7%!qgW9)f zDmBTTb&()8hV#sH%@pC6@0aYAB$imy;-_h-&<>-_sipRQR&`yZ-xuX!t1FEUk+;&u zPIV2;-y1t16>l7#g?_%Q_veH>?3`lS=<)C6_TKEWTB!DV`s3KpC&Fus^v&U0)#~2n zI_Zb^EFM9(7XGGuf7?p3@mzg6{g!Ie)8d66;xfG$7y?A;IlbF~=}=E#0^JZba5^81 zI15pTNJ?Iw6!+SgFQdy7w3lL}lypTgk{WEDPl1Bc&e;Q17Q!ADYcLb5ma#6)BHz~; zf$=!t>v{DW_LlYaDqB^xHiS<<$qE%> zEmOwL>>|B~3o~UJ+B{?e2}GPdrW|rETuMT4o1={FvqZNQnP^%0>x?RZ1*!9HsvnXk zEAR3+zI*)4CjjSb2+2^U=2{VRxRM+LX3_nnK2-7mYrW z5jXi}*^6CJe$?c>^v;725NErN4b#)pj3@-BW321=eis=d8ux_7>m2rVItfT5YN>1pxo;M82%iZKzQAJGMl3=3Xy$yTt7kV+hRzPAQa zLW8IJyPcX^>kS$knOD$V6|1Q9wVFn=twF;n^J+6U6Qj90~OPl2_%d zIC44Qkq)~3M&$dVrU^Q)by%W}X85pUUp?)u?|t3I0AJW95rdd#@$(o=k9BuEK;jC( zpcW#JaQr<(fJn7gn4kqo>n97Mfv*?ROMAAhAnm~6(6L`&A(2fYS9Erf)>Szm%<0NPP&0!lTH8@mO)zf6R!)gUUy5FV;6DIXs2~6=3`}Oa zutj#1E#PKJgAXdaTzhV84q9_ysZB@&ak0MH{L-}N^hL?#N?CkAc^(U4$Ychf#)K$z zF{2GP(@F({`%TP?tlHX-4Epd)73;EIryA@-gEOH}k! zWH2;Nh=ko+W@}=@wE=k6vKKC=86ZK8d$6_yDeQh`-9V8enG7OlEeyw(C`Z=RKC{OD*L$$Psf5dj{pb z&n2cX$>K6cx7Yal!&K3c;_nfXbg_60p?RF*$tq#bvt{{I@)XNj|l$ zcaQ5cUS69h4_IL|_mEVmrN{}#rfquW`tlJzd+0LRV9^NCv~b_bio%-;s9ilBIu56_GQ9d`nc{al>qhuMzs@u8oNe%Y0_ zKnsiG9o#vJ!6R^(U;=-#=5VnP+O|X=?67M_og;pKAW#$6UPXm=8^5^orj|=rqjG%{ zKl|DK=bO9s$9+A{ml_`15-0T6HDP$j3R{_eWT_K0r!>%vLe@JlXF9~ih^axA*CBa(6 zzcT25kf52Tk)5Nhnaf}5OVyCE$CE(&F99@T%q!S55#NqV4vIIH$vk`mz0e0QY$%~2 zu5sw9Zq7{7D8RGXAF~>(z{)u*1(mqEm_mi!;%;hY?yKMBiTVLi+w>>O?;_A07#n2D z@x~BJ_q*I*B)@r{>sRs3Q{JarWB(354OiPVR@YSl;t90XPPnSy#jo0=z9iu_FqYIZ z(newJe#A|SX!?GI>~JMq&(uk|XysaYV!>#m?Fn;*+&-IL(kB4JPQX`ClM0HpiYrS5 zpwm>zq{@HaO(07x`G)&85N`T>oYr&oxDC@2>6?b5iou|yZttKAkf%F19g2`SkYdPC zBR(#3)wV$s3yPI6mAhr*Ivj5SdCaKIg^d1f*4Y!TmD5id=mRex9*lPh=|X^5yTj>f z++PY4ZNtLG1lEBE7%rAZz!yXGNgx7tdt|>iRWR2riaty_OwFbp@&2_&gRAwt;Tz zQ#jN2RH2s~f(dP$oyB}x2*4n#0pMMQTzf;tB01ao#Nu;lG9&T3hqD(a7K+U$TAF8y zWMQBURsxJn(V~BNMo-e&MKXu-NO5snITe@P+!24Gp+9(~Yh3Hh!7YV`h&ko)Q9DY5 zTShNT7)N_dsYslycM``qXoYI?pTKZ7!uuZTTw=>W&OVlMsxjEcKDmpx;|7a`J(qb5n1NUzGnM zCbk!pNZxP3jbw0J*@Jg%ppK{PjyZrWS3U5#PJPzO`#!*8EfxjBxf zRwJmZ#t;IV4L6j;ZEShTAbl5*GM|!`(9!MC3TF6Ja$Y* zPFTY01ss8B&R)W^YE%M5UUq^p2MQwU+;v`DaO~ z=S8XgvW7P~sDj-_+I5Ip4>w0=PSWc9&K%9@XG*u(kNGb$^8X)s7?< zvZ|HUe7`u61%tsHA<)vi!$|t=*wkTcqO4$u2*{cA2yX@$$J9<4Dyed-TUPqd{@96r z(>8-BrT0;bzWLk?_+WBgOYr%+;#@(Ak#qD6*+`T}ca*Zjk+enPM@Zm6B&}>!tx^9p zA~o%p-um4@qn#W^Q(ea!zHkU9N6Z}yYT`?D&(Y6aY8NVZM=z!3Cy3=px5wCi2?~j6 zu$ldVk*k4<7ezkA#oIYGMF|XWg8aL>?z9*zacseeNq8l|4`vrxa|>=agVQ{)-S;!{ zngk_tUcJdEdxZI!rH>R_<2&d`=BQ*FMK~Si#a@Pw(;MF~gNcMgfmNo3LFYI41V7ac zxeu;Mwy)jKt0V&LMoUGb6^@2v6ED9F&4-Y*y%;Mo-5;(1;-80uAVU#>GWU5YIs6x^ zt@m?-Ceu_9%W7o4*;qVz@}GZB4!Qu>j?Z_qyJ2>krp3JP99Hp0y`o2aREQ>-A@POT zav#>$S!2phH*tm9Hpb&pXOcRZ%_P@#Ae>{Jb6os1lp;Ov(x||IudC%4DF24BmuSCR zgYnBl#ZJXPLyM?s{hl&RR+_H(*QVm0Eb(Zpw_w zReK)ia&DQcU!eamPv2kWVHtwFk%j>QfkXuX!Tul2W64TyWN%7uCYwlV#Q~_e+Y>!rCPW}NXymKwWUL`g z8f^F7<{Z?xF#?6_c3psMGzUYXb+sh(MiCE+jJEyZXEh4$g+f%5cz_y;ribDSWeZ6- zHY0J1;6_Fs`wJEeS1@pJN(JpY@;}^veN>S=Wh}dn zD|s1>-59IMZP{m+b-3kw1*G3M10hVpPq}kYVL}PwZrO;RpEiQ;j~aQLf>6Sw z0|`%f_DMNtyHNs#35d@tZkdYsfql>Lalf&|14}0lh=E4D%ed_chB|PdYDfk+4?w_5 zNGbimW!c+$mdB1#5kff{;7{akCGrkAcn&5Sc8fy|tf{eW|0CY0~Z;^e+?h)8_lR`n|99 zmg4CM+-xu?zK8GDm56nNDoNQmXyn+?Hy6Ao|TYNAuaP zzDRGe&7&CmU}pr}yWr@<*4g$_1(0l&O{U7^v#i4x@Px^d`abc*o2NX0_*k)_K=@Gp zagMY+sqhmNem!On#%t&(tgrXW&x1z_{4L$a3(ZfY0b;63G`YbkUF4o8P@OzNrSel5 zi8C|I;vH!z{KYtNSSa)v<0qdq5*IQkDBw%ZQJ02#YQH2&K<5G`RWa-)O5Zy?hKU9=x1IL;!%rACWH+hoy3)mm0WZWJ< zP;vnzuCq4{ zM!C)+%k$OpAvRpL4Bet3R=d{8!|blZtSE{cM8!>8gzVJzyq#9`)W1?7BPKjY9*=ce zs7>nC+)_1v%`X%D?iVAK@QA&V@RG_n7q#^LA?LM*^}QJ62fB?EBKH@8h zDY9={jq5q>tPnu5%-A0rmR^-NM-Fbsbq`o6>rYcU)O;y+jH^4rvmYDoMy8|fWFzyb zsA{6kk<{a0AD#<`8g)&Cta1$ulz z`926GV5@~LG6|W1fv!)C;qFBN# z4EJ+Bds!NhLEmpDCS;up??Y!;FAn!%-^5eky}>odmQx z>w6L}g$dWa5a*AeKQ>gDsu+lyw;T{Wd;66jdUE~cee3;ptAj3l%xK<}DU3d@yXMZ( z15>u#`j$tC*`&PgJp44_tmLwsK@RV~L{}P;qL7Ezu-)_~V1Iper`~-8|1}mz!SLn4 zz|-Yz@y?rFI5y}asX@4K{^im;c=r5SR5r4exIcU9@DUY%Rjs?Avic{!skNo%wXSs@ zIp%b0WoP}=>nAV0MoD&>#krVLWSC}3jIo~T{hn>NYMEHOW^LtRe|OipfyHXiJ1q0N zH8ue%*9H=TURe(|gssBzw8t(S7&)UJsc zbEr!b7Do4t@~|rrpVdcH(;7okQRjQH zPqmP_DPfs)i?6Oq6~<2bfN@r*9&SE9##_8;CnbP&!^Xj?!9NYi#NP(wrF2-*6dVL3 z@UH`i@V5a`784Sc6Elz&`rC8(sp>i4NC5nfHL4dZRLJd%E0qZ<%M(c;7b|Nd{VEf+ z*s08&>=ZX!yIF?=)37WMe?j5I?%$X-0s9BihmJ;c&E37u49d+l@hB7DS#Z8j@jY$$ zOy^vuP%6=5F74e$Cfo1^5C}R)9l)J(%#jv5B(@~FACe#5^WD&n$MiBXe4ss>?6w+! z!^?cGnA>Uz)qB3geoK;+m2#p3*M?c*bA7jkFSz!{Az2k_69S6nJQKWPa7eeY3j)oG zSO!3a<+S`u#J=>^L;@;V){7l^XSX{y$r$lSm`4T+*WV)-a=Ts4lI+OrGI+ygjz4Gy z*4KF@?tX%~XDQl%O~4f(vVm-X$4 z*f1AU%Q4p!mloW__zRJIpF-wo8-GE!sGte0y0pF^RZOm9I}uY%lVzxN>Eujv_AvVk z^W9(U#^fnQP=D=vE$n6?#1rxgoxpN@8QZ z+dpSPn?J?p?k+qj?bAr)(6S;bJ~5@nQ_>jBj8-HxNb|B*_+2%LlO?f&=JaZl7f%jd z$Wm8_(K?F>2)@E9*oQ4VW$fO2^jG?zt+l4DY^*2vUOiO2+d+_vS-eh_gtRX2c~5rq z?0s;6BOHe`PjzVJ%7+$9%$Ba^RE4bo>S}-NLL1xN6<^Ibm($O71QVtk@ok)TZie7=fjS z`G7%o^ALw{bE(tE{jIlPm!orPsJ%7O@V0|kZ^y~^()dJH?}Ed%9|3c{Mphu@_QW7J zsPoF7`T9$Vs8k>QzSh&8oT3WhIX}>m2An3&>4-uH-Lg?>RjSO?n}oMmgs3_UykG7I z&}8s}uSB*Ngz;jkFZ%(bDSEFH9mM9$xCPn1HOw>y+l$hIryWk=(kp&u_YhNny7T+) zvHB1|WV&w3HlhIsaP* z9F))#*=0gOZ}ba+3_58Wpsz+A>c zyqLdB&;>?;7)dq|vKWr($=1if@^|d;@3Fq$;W(!mTk*6-e_TB~lE&8>k20jp4La0i z1#r|;Qwv_fz^S~(8D-&J_J!f~dk!>1phxV}pstbXMlA7M?c{#a01*21`1s~s>~|PY zRo2iQv7xRjQWfRRr?_*hUgfclj}M5O1s^AmW&GB}D~6N}0qFbAD7HA7T{UU-9*X6sqD zYta#;Kis9WX2xA)_{bt~6^uEaP=Q>{dTn?!4>r&-?!hc0=X%3s&>^I+gkdV*{;HR0 zfKG;KT1O{iX}G4=oL~q)3q)gSpqnu5*!iUSg=ot?vV(Oqfz}K()q`>3413}UZ(HZu zpNYpxXj1PJF3o;JZ^*>ayXXwmWWRdaF@*6p6TcfYuL{nYX$As9V4LraGD zZab6qtc2Q5j_PMYNI5fc`guF_dyMBcH3p5jcALGvl#jf=d2o+)lu+WC)t*>lrJb+l zEW+?1(wzHuLv9}P&lDq<&h~c@iXRE*B?o#`A3u^VLIwr)Y>qz79Sv9v!*YNY-({O8lA;|F)JE1@&8=6QfII zw5dVe$i6Yur-V2-kor4jF;^>YLJemv<+X4(t>duaY|K2rjq2`V+Tp`}8KX?TEe`vE zgFLpfEjqJJo9T7aCR;jwLCLSaA=ka^X@Oe(kC4MajW5ma&&1wruJT8x82^V>L0S~8 z#^~dXSyb|onsGIq`V z0`9MHUMug8Wu*?{ zEx2eP3Qs#lu3Ig0ACeY*bk>Gc$NY}bj>i~A&?v^-jkbV|bGF^Nn|*rNq?1}r*deFJ z%Er=V(rhOh&6NPwRjeuUk~@0&L@og%1>T??@kMf(T}Y&@mqA zdE@X9c<+enlio2y(gyv^+%XvOx4kCZF4eco^!9seqiUwx>V?HUYIa9H@dt#b`=nTf zfkJuBtChY(es1``2ai;|BhOis{xcd@`sNpca@U4D%P`@Hr~@XvTrbHL7irR>HKZ&! z#RS@NI>_$!Q>zV1G>>%7VK$9eo`HWBiFOFx%Z;7o&b}tsgCLw%Iyi)lX4Magsc!?1 zu*b_u@wJG|i8)5>1=-nsh%F1dTSIKU*PAYU`xjtm$qvp5xNWjJY3QXU-MSl}QWVQ7 zW@U=2?>bx~To}nRB#DwFNcKy}oLK^x*F$ok(u#j1%@iMl6`DgS3qi9}?|#I~FzHJd z2?ty&tI9I=cebF5>jwAdqD^V&z?C4N<NuP-^n%{r z{0og(h(UrlOUm%O+WAUf9S%CIn!?hlS=l=R=Ig+eAZ!%yl~0S>Xif%MUchg*+rUy!Kqqq_9Bnn9fSyU#_=C{oLhKSS5~f;nE*)ObZDNuI=8soSIlHq2?*f2U&GdGNOc z_4BK=*4SE9W&bb#{Jx@uZZ+o_g5%C61?f@W+Hr#~n`-)p*tzP+T>Pdl)ftJXDMa6A zX(vk*59Y^U-iZbesPL=NSJ-iXxFNte2zg|@BDX*hyNV+vjMig{_T-fIz2{Uzi{U_m z6bOarKJue?xKnSC}LN5$w(8tZSBV65BSeEe7}%i=-pbZqyBPD8%7w@+;> z)~qDjD_HxdOa5-xFGUUD&EBA360pG&3k zY^cDGG%(=51NAFZ@76!nQldhf;Kho~JoKZeBnN&`KxSv|DLqaGWdQE(1{M63lt(|f zBz7wKqnEBWgLjp5^zPx}4i!lmM8w?D%QT^3N)FBPEZFpkHF)u68H~0+k4vG`@;sKt zC^3K`ePuedLM0?eGBqkUb@nN)iy^NI7%U5_T*`{YN0M?7R(Hpzvi#RvWZN5-2zQOAL(Eg zPuWd`h67_|u^+Xw1)b?{(zbQjUJd3l@_pZ=yCeUDug7(jhOf`h7wDfSlO;=DUS&{I zHJ}f`K7MWy;VTmXU*kU=)6_{n!Sj6HQO((lQXCX&&z!I-L3{&hXV~FQ?F@hl1q%mx z!5XBB?KBb<1;vxuh8DTj70|fPf%j8CC7ZolTqWjS;;YcS)=W^8sI&D={Gc$f4=3C; z6u9_NuewZ+bl@+vnMKTf*a2f!+uU`-UhXW0S8PG1ZK@?B6GNFIVEC(M00=x8wDK|e z$+le>8Dgvusnlz!3suCjd+1rC{y59JGxywO_WHypm1V{hZ{HO7?1LBJBUf8=-L|P9 zjKDYSiLN@q_sXW!(@iCrb3T#e9I*)CoW0lYlLpUTEWOI&WV?-vtTp|t601GaVR zwl}K-!pGTlzx-W#6SwmSmSVnqS%=bo zK@6t;z>R5^l9V)rcG& ztYEp!Q((Jj`n=vdvf-{|n@HTn8wp`Gy!&4RATna(LuZWYohaqnT_fimDN&Kp#K`M$ z7`+Kh_bT<`v$J8eupKkS6+rQ+thkH^^j}rgkVyJouqGOpYA|p}e<)}NaMfjU>rrDL z93JxKeLK>xF#oK|Pjj$!?%bgJgjbXS1p|Qm{}Z}@g~7j`|4&}`-yr`f==Bf6=U?gn zzd-(H?CalP|H&}_2MqkLh5irNzwGmWL;NS_;~xk=^nXMAKR(F6VgA$o#Xm5gxc~Du opF literal 0 HcmV?d00001 diff --git a/H3-Continuum-Skill-v1.zip b/H3-Continuum-Skill-v1.zip new file mode 100644 index 0000000000000000000000000000000000000000..f24d4810ee4f8e9f5b1bd64f484fc90f225d0b2f GIT binary patch literal 17773 zcmb5!Ly#_9)FtS&oq5x?ZTn5zwr$(CZQHhObWmrHd(@vAvy(rJbv*Eu94mor9CTt%Hj* zgR-=wj10Z4Nvg(n+;JPS@47nL0S2VSWcn=m7~CV{Mops7U)Y|b>)|26X0obmyT}yD zD(@}*d}Y;?uJH$K=mRz{(k6{`<#Ghob=Q+wi@S4N-8NSoA2+|e@z49JJR{8VoT|Ap z_AaXKgS&IS`t5{^nG_B*V%NEAUx~Cn$30);ybFT*^I4;?y-j_yD{blhY9{V(pMv_` z<{S2BhV+%OWSpCWp3+ywFOpG4<1arEk^8SMM8}f98PnMp7KVPz^iA0m>zjh! zSGKG{gr?M@4< z8E%adSewNMrnVp&CS={+Klh$L+&?Y52kTd#(eu;u97-b$?}((c@37!Du>@}wl~o)> z;@hX)RajXhv^R1PCHQ<@)%d=j2$tu5PGz5u{x~214oZlQYxbhBoGZ659RB5g zeK!4vz7X7@`i8e#bxa?x%n?x#kFlt0eu5zdsnLU?WGtO_(j5FBc(azF+J|?Xlkvpr zn>jNRkc*!6xn1hI3+q72pIH1;Tm_54um&(ABn{8YS;mFA7J%^a(ql_Aq_a~{l-%nx z%UVk8^#Fp|2Jp24CM-$8NIaiW1{b(xUfH0exc%^1S*Sdy({LJqok-6#$8>;n>4n5? z(pX`J)%-h^`ISWjiDxFigR^JAb9TdE!L_pxatj{9Lh1gEGMSc!FBA>ZP=aOW-|I$` zsB~g7o>!96kuSnS&_Q0ut5T_TJQ@lbghhtp5|M+`QKEJ3>1zSqdmKUWW&(GabQ~0G zDGWNACd4j3fGH#{wml1)T*GH6q^`=j?1 zF8P~5j~2w*%qPy?W|?oed}|HniV&ofRd`g9J1x#;xf&N2L89WM9=>9I9-f8axA@qG zw=bmlAEY@|aAMEsu3Jvwhp86Nu-q%rgkGEzJg|3!-OFxK)Mf&K?Esrbf=vhf`{e`e zJmA}rgy9sB6~6|k^f>+;zAhmM;L|k`?!A`xb0sGwyIgdq4q)osl%Z$7Pj))WE_E3G zZc@Ja;24)4*HV@Tw7ZV+;1=-ZfFb#E!YtrU`znS_KGvCVoZ|y^QEI=>XYNRWZtBPr zR~9#arTs+^8R`anBdmI-lusC^se@eno1#`od=HoPNg>fy zDGhyS4sbj1*);0@GI<7mR(&S7F&0wcxkY-m)?^B%=3;SXll(YiK;NP^&{V~U_S3}} zi40eqx89d#WYj{qqWOdU)ns8*-17t2+{X;?yl^tGL2?w)PKxCoVgFJyD<2^&J7{c{ zrXwJAZ-P^(oN?&i&k4VB?VnmZ6CU|!$riDyUy&wuW)@rR^GV5?uXx~lR1-0_mT z%qocv{I!5 zqk;G2)l$u9=xkQCrUHF_lo&(UBwg1#xOrj(7_WIr)j?oJ+;-3wOZX z-LU+5+XikqrQ4<;64iMo)c{m8jjiCoVo8G$N8DzpZ*3kv)gwqtVr8K6JQ}-hLj^va zKyF$Rw{$#xv1!Szy%e1V4`Rc``bx|Tv5UVoo5pSl=qQ|%zGql*PdkUwbCg(K&otNU zF2_OlLt;UHfZbPhN+$<(o7KsdBl(xufS{n5o$46%Vu+SQR1wTN6Qvxv@VK>A;U!`Dc(-^>U9wI-O;5dGd4l$35 zOvKe{U@OI@F`k6~7E_Pk9mgrE3T*HKI!TOnj#ZZ&GmnXhgL1KM5lh^nJ)?0XEaKr( z-7>mgQujvDSE=-V)9-r}v;zrrbN`O|lLs!14& zptjH1lmq80jg%t2el#{*Td<0FvuQNZkc$>oSBugenoSJ1qY88ls;CK>3gqT z^xbuJ${|?7qH3#ad6jJP}OD3Si8|Kenh4HH`0=K3f&(R_4xaim@6TW~&zUS$cqVe#-D* zpFSnIzE`5^pOEN#Iz;Rtp*QN5s?4xThH;AAa!P={M2}**PoYxz(YdV@wdfz6e+_c( zb2P!yO`q@FrIS?@@(co-HEQ|>p@>&zQ*agpjgkmBtU-hR)TX^xiIqYo0U+zoMDFam z6?->4W{fwqNIJ%W^E3m0c-b}i z?v2bsITNDpA{T%afRSsybvSaZx+eE4S!0dwY(Si#??)!Qk+`Hrn%U%qB<~lCW2Dxv z;_2Nr6o%L)5>DgA(%ea!?|cPbAY0gjRY?+2KiN;=Z=XP@oX<;USE`tVx#&_bh2QUkOXW#H|v6 zaA8~HD?A;+Eatd$CgwtrXVQ)s(;z3jIs8)<;T52`~kMWwCiu zc#&jrO(5~oc*OPLfhj<5lc+Brd~IwNHM$uKF%xVtHPSag-e8>$IvB+4#l211=|$<; z@}>uTPnxPkQg)Hd16#mUK|5U;j~Ci#f8mTn3>^%mi9E>kd3P)9lT&jMf(&u*g=Az< z2^2c`uXRCzDZhSLuWfu9wQ8?^%@ek!&Y0>?{*BJ zze2m3-BKvcd#?mpOqf2(mi-W)Hyn-Yvy*0^)0X6&D$m4qi1mS>j~%_evxSOQ-(LO_ zs1Zm3w=}edp1}xmvg66-#EXw0KNnjj{OP+IssN|;^%{^g&SC9YWHNGN~2`?LfcE3g`udSG^YSYnI$vZp&)QJ_XzJ zF9#F85c?ucZt?LPht<#s!-!0Aq3jF@%9!6xdh$TkI>8PwHn0!j7Ow>t)r}$7GLr~W zO4b}}&r$j0Tgzh#fl;c+CW|9>xdDGd;9HJMy5(?};%On0rIV6<>8mBH8MF`Ws5TC3 z$8u9qL5}AYK{Kwy4l!o0p)XXaTxj1GfDu@zSLCbRt%um84YHKPQ=G~9h{du6T`MNu zshH~5s2Z6QGG%H5I(mQ*xkN~*s*|7!zEu(mbfVLMJZ9uAY;XvN^nkfyQ$U?ifD588 zB)ZSxEtyIL{uoKxz7hKY?@%`a&?gMN{4@4eZ5H8D?yRcWB8@p*DcYl9 zf`s#gY0meVvnsu`YEsfCT$4K->W~S+gT>1kL;s)y{ z+_4AA?C1XBvw`lD5M zUO}((-m11G^-h_b8B)HyY4m%uQKQw8HJRcIicT_3{OfN{Q>L+!#kd)PB?~mSs?uc>mUIZmW{EMX8vx#Xy5nw!+;;D+-88_D@-V1oR zIa6bT(hl_GXaZCTF(H%1PBS+rRy|rV3F>?k4}CgUZY$$tp*0H)24@RBrdmkFY6ebq z$#op*PHSr^9@VTBk77y`2Nik+I(K2JiyNfETIOkiqwRj;UCe;|z0SoVk=I8qm)9#H znM;>45uHx%`yc60xUmLO5Ae$NEsG5}s~#YWva7>u(ScK2~6$cI2M-KlhI5PZ@eP?c+y|s3+?<3WhlC`xDtmt=8L- zH0>vR*~_mNidXV?zm2P>KnOcv{>!09vIrdHJ1*%^^YEX)VM&%pvP_9M1a0{%XVVDR z=<=1nP48fx++VFR4=;sjWqhMOO+P@s^fm$uA`N)(40nBi-E*OTrj!c?4TaaM6e5ji zaO5t{C%cYbfC%p(^aF9czh#Fb;5}_FqQ@T5(>Rz%4n&qzM=yEs#t3o7cOC}ee%WCH`sMG|VO>iWK_Q#jmgvFftxt0uDH`Z&}LMq-D6 zVh;QBmS#2fyjxYeRx8rhA#O`540`_PGSagLb&KT*8yd05^?7%nR5ewpG*Wh9wz=i` zX$yqHG58Vny*Q`(~RHKkfJj>(p8Ei?fS1YkZ42#A^oh7ORiUkko(RL71Vz zz`f%iJ|a72E$lpf+}EY3Q>_8g zav<`DvG7+V4h1h$tin&e7_u6*F|~M|bsI?zz7Z{`pMm-}A#}#Lu@BFbE!6qVtB^sy z_n_ul>McyQP(*fUZ0V2-1ZVFg z4AlcoceJD!VCOU4UF!1QA#Fh5acm8uQ*)znFiDXLRyjp&TgK$30hVtn4!;}O-f_VH zB3me`ivWE4L6J*@TyWxze`dv(Qms}}J&aB}>C$%l9z4KsRH-Ck4zS@A?56lETofG* z#M2o(<7y3{GvM-Dl6eq7)kRg1k@2&Gj4h~ZVNCqS2SVEs7zxXd-k&!xMjTqNw2mdx z?zKiMzP?{h(D=fvi{&pC#ud1a{`Fgo;kx6XE@RAJ4}(oeQGmvxnHP$;Y%F4g+_bnH zK~<#=RRiSkandV%;^t(7^uD)%OSPO*>YNMF#ETXDi0pIB(8k1t742^2=??`E7Pde7 z|NX{)0PE`P_eZOF7%#7K1-*rdYRa7>yk;#S4-{!EbliVGo*w(j+)=RXG~~#dA-^$T zPzr7?*)UTa>MQ#Paf{WU6U~?#XVpVRsL$jnc;wU>P^|DdEnAsfWnh;P2O9@0SM9SE zS}W7=pT>^w=PQZ+`#gJ!^Iii*8-%gPW+@rK-jTiti2sS`1W**{bhRiD0olOyEPIce;vytbLpt)a|T3ik?P zVajb+@ILlu{oGssb^j!7|Embxe5Wzn6tkqihdTAmvpcAhGI)l)W>yUIjv{rFag6Ff zmtww|kVON=8h!RPpqo@LQymV4F2MyIrr~dJ=A3|u(PTmk%1lXctw+6%@x_FNo|g-9 zIXDqg7h?7>kXZsrtDPE8AG z4$^GAcZ!ADEW;}KiQ^U4REP0g0!k!r7i~-irVz!wo)irlf6wt3ck#HP=z0VQ1id~- zpTyGID>}uB3U@ss)dKSiy~X|Ra}JGFa!rQp^W^pvn_m0WEcCNULQ}=~Mxj`YNP9j| zC;Ql#{XIjQWD3O0A%6Bx3wB=Em(1^a!cg9jEJ`{h@AbC@Q2z!tT0Tw(YG(>ruEayv z0D%#;IXfDTg|1RzpA@xLGKpKLVygxue{vs?Yh@dcI6 zV2#>5JOjELPQ@opbT-1)bO(EZLZErhs1l}7e!e$4%4!oN(epy|&!Fqdpzk@8378&n zUNX`T61zc_L>JWugTmFa!_TyO5%{(*`t?9ibSB&*zuV385d_eHu}n7_%)=EN5%oQg zr?R_P^{GZeUJTK}_Xyd`lSR}LJP=?wqh$U#=b%Rz_E&{Q2l3YixN+h`9*PVm(2U;E z0Rt!?9t7iBv_@%*hNzR0gt^{5w5&~tVff_7vZu^S6cUI#$hn3CqoknTdGh0e5r_sD z+<@t{e;JgBH3&PxGMj%af0b)wbV5KebF}V3Q?;wh)jb}vDP?}n$8(*3ZsI6V^CZ4f z7=x`qIhv%<4C%)m6|i^@sL0zSAHow;r(5!e3eq#GKXh^DHU#OeL1E>w7kvmn2*_I^ z=Q60@uJrjW=LX z6Q*Xe?_h!GKlxs-J@AmY7T)ZI?llUv*q#H@8M-H7MZ73p?T{E2xN=?Y0gz~iASDOV z!(P-Kofv`k)qQI73>f3s-eJ1yI1>VV1`3eiLVKT7=$k^cpF_8NDLa3#+9iOhQsKQG>W%`Yh@>T zCMV+T%&tHG7PG^pJ@Eky2DneYIXWAoepZ;n@S1TKw}0o=DcpU!&!xa`((|w~A%>a( zc{$O~g%;e=+@U3qssDW_8#ASY6e3PjQb%`IJe}2`Q9n# zbOeJAnSh)&c;`=M^pVgoQKIq6KyuYXV>F2B%N*={i3RC;zz#T(+Xvaen|R zgW1O-kv}W#D6>920>-M1qe}<%KEvb5Wr3C~0micElw`O6_?SN*&xeoet4It$Ik^ii z%n|mVaOPn-Vm6a}(`^2^qdQ#ZQ}SN|kf&rHxqlJEWMx$zta^7E_^}X(JN@ozxm+$coBWDw zb{qeOSMf}49k%&jlgC_VTNxpWZVu|qZdTP+`Add+znq)a=2E#bmHw9Z%(yCtEj7nF zJ6S0W@}*+(s#M3{0vE<_t6~u)zu0g;OIcma7wY#aN#~cMQd=Lqk41X06KnSzd2;!@ z3j6C^s`{1C&%6GWkIbmDdsnP-52GV0D_ugV5vI(Eq)QwE>fDw5<5}w$pY65oh<=8; zM!x=9Y}8|J`uu!dl98#kt;%E6-|^xfaVVos9TW&uZHi`SmsfYytE5_=%H8gz5#o)o zIfv4{9XrR3r}oS6t%|lX4q*$NplRS~EbuGq)aN-Q>@Kyg5v8gJX!)wx%4D9Zj#YR9 zS@02&0+b0r>mwwGdVRaNm@Zc9s!NKLF~+wt6P7M99=<=iB%aMyRfze;Di!|rB6s7` z+yhISoiFUlbbjI`3=Tjui+xb*6I0cY(r-kL=xEK*8V=-r4}(XC7D;45S<7Mvvmp z4|k;!LGG>%s+>AVFo>bu{X4gXK@U+)s=bWtqK%A>eNt-QO>MkNVv<4mm+9*Tq=*fX^DBQXkf2HqFU5lA<5?#WK zMKsOJ<1~>k@@`cJ2Fdy{&0Ztt?<*Ixs-?9!C_rQk=A&K)J#D$GzeH(HPy&^#cf2 zV73NXEJPPlKhUxuLalWV6dO;fd$80Xf{zf`U`V--wG|}ao#3Q{o$h#mFKs)O-Yr`Y zTDZXeSmMlHpxVIB$_zYueCi9&D4NP`Cb_`wfhOQMv7XJ#m@^KlP-vBK0A#B04>0}n zKFGnq!&Voo`$&B(v~nv;7%VN*ej}E>to0&Jtht&IO?e?9qVm0cdCK4Pb(JWnaZH7x zA#D}4_~cp}QZV4!q?f-O0PFA)_3T6}g> zGe#roMcG7XLtk5$19dp+s6njM&=gv+$L_{sqjwoMH>G#B5medyY{3X)3=w{W1G#bTOUOXd$6G#O0=@HlAMxt&7LWt{WeE-kJ?RkcLKyMK)a z7#!*x@D|ES1dNbYkFJti1_6%hywcpf;hr0x8NfQ3@Q!vL8isV9JZdS0^lO3VkU*7} z3NvJpLHc5-=Mqbo$imFNT0Bz?N}Wl@=vD*JnIa90lC7Wjzx!>%PovZXd;b81AjDYU zFItcZ9o8*0B@M!87VRSWL!}UdsQf7gJec!9CUhqbvs0_$B%L0Hzzb!6csse#jB2h5 zK@jwAto$oh_q3l0Ea{;-mf+tqHqavo4tyQGe*vtnX(}$i=Y^#5_FGF{tsZW*eI2?B zw#W(IZ}a(dni|kDJ+dz-q9q-qoj1AIl8VHu&+3%I>;~?zPKt;~`7soEdV3Cxeh4>GZvA zd&7YI*dDPbEXSo7W7e|_%e zf|Aqi`TO0UjHvxsuAiV-LwLKlB)6-cN_$Anb%d-0Jy)9PgW{Q{5uP=| zk6M>JDvq5%k2~afmLrR_3g{V5FDa_BFa{rglk9MuZ*jLbuqWA0R3!7Yu#CpDaLEnB zeVU-&G6unKI1v+sGG|34J2H;ZMk|jK#~%71rIpA={&iC+A{J$NrBsKB0$XmK>l2CK z^4{HT2XyPqQ#{<+9(xTprpxPZ2Ic;F=OrsHtIT)7*Nlw7$Pz8lV$uXCI4QqexuSrA<>kDVSvor8Gk^J2nX;nDNqIM?z zh$4(TVjs)8B@(p>yhnYM`t@iChi|3J8u9}2a;O7jyZnKeC^0=TQ~l^md352#*iHm5 zF3@xxq5olYV&!}tXl-TD3F%m9b`%{-Ho~%qWveLFNE%6f|MMJ>FuhE!1&&2pm`@#J zX^FNj^LnUtH2fU80 zaUbAbm9Mc__Ak>N^+;f5nw9d+)r&p2xmctk7CufA*INpeONB@ex|VRVAWOi7wc;}w zIHZ)}k#WkNWSf$tvOk5R;N#iJ4Z-}Qf>9#UX@W$(LbA29kDa@KYswh6IhcV0Lr8d( zV5(-OFvY=SIsC5%=W)&w-PyFM1!|}>ecjt)rRfG`Kwc4dx%3(E9li>0%Llpqn~>MA z7d+%J$#;+Qds4^Xi_@vNe#8Nyd7_I#hG!BF#G`oXN|dlUDx3TKhq*2-zZqCS*mFWHR65x<_&co;9rsUKCiVi=+WLzdL1_Rpo6u}Ed+_KIwQ8=kZ z4_Ol?T#zM)HJ%HmQvNk4$yclFzTt0E*Sy9NoW4^SjYuodZNkF+MAb`b0sv#80*}F~iFTK}!i&3M2X46MWX0eQJDsz|FZfYF^%Yx< zxTZ+|NKXdKH30ijRXMoZCxHrZ?E;I*Ye9O&AraM2!f<|7@0t{r%mwYAtf_;GGqWHa z)lZ&`|H-ZX!k$=Kpwg#Ml9Cv>9L<=jYv;+rHC>EH&&r{Fw(RtFi2eusKduy|pw|7i z7{adkuSCW0U$2Vo|JarPRjUl`Oy~?EsrX141 zHfyam1x`)!tq77v4iUg;mZ+Q*(Du51*O|^j;+edz%jy8Yeth)H@xCq8i_pgqzR-YJ zCOz7Xgz8!n>Q({}<)SpHj>U`Tx0AzsCl@`}u`0q=nk19nAormV#IO(?CvUWfOWN=w z&r#GMbV2L2n)cJ%wpM6aKj>sGNznWaZgeBw^V~yz)~|(bps#Vy4rV!;3fCepG*0mj zL=?Vr*@(8d9J5li(|%*biq-2uBIsFi)0tc{@ZV#@?j469NO45*hJk*wrj4d3ip9Q- zn6jw02t|l^v|`n&UzZdpO>~sB50cT^U9Xje*e>Z6klILn_i7J4D&HTGpO9IQjLwau zBThD*s}RocTeg}jXZ9*irkeCm|7&uL9NzNE|G8GFF(s}adDX>;#&+khXDBID`Ijsi zxv%r?W}wzW%0?)cQ*lqTr1Tk>O>l*vd>u84{H)+PrHt+cnJhHq(4-}?h14=Drg3?Bv#d4CJdvW9a1Gba#kyzqH{o5vuWx0 z^l*E+J5OvC-O$*@FAk^57mTsTruJ8K0KHUptksq-$R4$J+X}i|F=kOWiOAF-8!0&e z3x1(+Br)K@VhPLLhcIxE7e{Y>)t12 z5+XxVXi@^xv}UZCM)x8(*V_*D$ubD=#LIWBbp*32Sy^!^4owYhWAqu02KKnhmf-uK z3=Xo@civi{SfPjG{}j0Kg+H?4P8Z}} zGE7Lka*dUTcT&etlOOrg_FyX&KtpHjHuM-YMnXI|ZnlSiuZ9jQ0!&(Q!<=lcM6e8# zJ}HnBK7>A6&yFp6KTM=FBkv@IfIb|2lkgxv@F$zgp`@oFh*mz!899R~(B71lfToCu zTgm<@1{tyXjLw`6Ehj)iPu#)UHe&5V~xT1UP(Vlgg00X0-aK1fLPpVIsSFo73p$42teZU4i}W) zucP>D?qY5zyF(o50Tf4eQi&6SRq82@hE2KljNMvB46jh{5U>-7IXNK7p*3i!qJ8Tw z76Yo`n9k)wk~7t)h!IHGRqE-YXqU6L;c^aDW-JB;K8I5!9etjdeoaR$!+5OmtgxT$ z&IR`arjkk?nag2HAT(Pnu@JH#AHw5d_t*eo3?^!b45|o}wAP9BAQ&minVE;Q@Wxy5|9Zv_%gsBIwzT@nH?BdQG7lp!oTl|>SOgbn3A|S)%+Ce_4R`Oya{5w zPwC#kHbZOm@7=6w;!qy!wP9IA?jO3h=mT?I>yQ(>Bm|u3hHq>sQ&OSF^fqeW)a@w4$Pd5i&PknIvq$pTJ8xH<5^dP@iUVWI?+x z_JPFzYBJt0qTh9{%jU?}=#jgOvI7_?;!$ys;`?7kEoFu%;_4m@muasrN(o3If!B$} z3uAO*Tc*s@7-Ch*iGGF`TB^qPP%oSmQKV~=3sVD-DO;eUR7Ve5r*9+0(V{3Ol|DMIHqe2OZFqP-V`Re67zUsbr}0U`bhzo0Gj~PLkl7jrcINWI|)afs(IQ zTPo&ubd1Y!Foq}}DZUrv#D^6rL<`6VuelF52f?@W2E9-dFbO)Aq%_RvilG3oF*dry ze;VI|+y@vi;yo06AsIwgE&_t*Q!Vqzs79LXNO&*MLz;1*WX1wXPrEI3Hq)Wp?^83$ zzOG2LF?7j!7dkKn{lfU|dcrhJ3CvN8Pk|+JLu@RPgrVK>m!%NoJbkhcu|?}%o3ZBY zTATY74aZ26lMbVNI=K+i`}-P#Asuo|koZ@^R=%1EX_Iet>%2HfkiDF^YmjPw7+Z}e#tGy{8iRWDp6#r_kUvX(# zYRhSzc}Yzp@#F0F9Xn<)lRiboB7z1GG{@f18kEw<&8IMi%rSk!sV3c_KudKrGc8qH zb*+KH{+!eBi2!GavA}px3MW2@k`c#xt2=>lVf^rs8AM^Z8$>oP!Iv(U&d|H>RL#)p z|9iNm2)Acwr(%1s;@%a^Y##*^%XAYkj;Xf~8`sBJ9{MzrI+LNum`jjSZbK^H9_=V@ z9XFV`yV$+sXm1y_b2-?bR3uORH)|N4Z&mi3J*Lqu=_Td!94P)z;Oyw({k{0>@nE8H z?_Z(7jYtP2dbXqH9YLAEN7t5rXCUBcP~-LHx`|tWCQXcshF&fDLx=vjNMlC*$%?&S zBe%NQvyoF@?F{mbeD?Pb$pOcv1V8{1_Y}3cgSD<)$V$ zf0Gu3(m!QEW^l*d>gva+5^X(M-TZPcifsv3)~9{pK)sNO7t#f{YEvQeEGO)IyyidA1n&ip*X@yX-$b<;xn zJEEpq^wW#UGr|Ax59hzLq6$f~hZ6!2P#zHw5Xb*9|NeKap3c$L(B^+vuyVC@Wo@5Z#2+U^MOA2Sr0|A-(vnJ``k<>{fBTHE!zlSZU52bjK%6UCe`4(+!g%&9?z zY?R|Bzn+xDRVD=DE+`27>Dn|+9W{*Mq)8_?)e9ZN4@3qB}4zsqw z`a3+)yfug$UKroLZTsi0f5R*I&-N4%=UfV_Mjuhl3IpeShE1+f%*JhRb+xI*_EblB-xvs{ z{-A|`hXIKOFP2By+D4TT3M<~4vGetm2k7w=Kh9Hr@$gg(UcJSRyX$W5$NlE&3;di; zY$=Og0Ii*k4V@jY`(fS{D}mP-z-G8xb{&P9P7fj+UNn*|*dCWynBc}nMQR|?+i~f} zN)-*f*(?-t2RPhu)!Y$zZS(W#FI~HM^1}y3^QMg&-Ubl4Z`@3c#@62M)SKH&m-em1Xl;T_X8dW=Ndj0$3M*3xf>o}c+SY2Fay;rfEcd3KG4bh zV**A6`jAX*NXtS@*FiUNM<7wKV_WMBnO7YxHHU`&(L}WK<5zk3aQm~gBIl_F*@ z>+cKTCEoL2Sh&~EAWh10e*fQ1`hSibYI?s3#$_Q)cYDmz()f zUA%GpxZu(?LbR+qbVySt`F^@tgT8duX4v!``3k;XQpu5D_!e4lxL>;!dz1s?c-nighKka?jJunk-(?8O)*`oVwPKVkq9Cw0^#*ES1Y zUg~B_@H?3oBH95ew2&qV(C^qL1hQMwGQp^!{Tk{u4Rnr%qoc6cOCy8(-niEs7UCr z@(O1}wM$E(ALXQxD33Ob9}FH?7Pk|=e|+vd8PoFm--iYHb05j~bZu@09bT{m3eX8| z=p&hMLE$@-OU>@(bIl>d`efG_31_8}ItjpuR?6coF3KvcLiVo4H4F+>_B_C>bD(UZTf0n~ zFe3#E$x*X0GPuQ=_~q(N<@Y5@B~}v}MN3p2no?LLqS8skMWRHZ5IAA9zWE}bX2^x3 zxW6`Dp;B5;e^C)j^7Z|sQm0sOYEs}Ww0)IOa|RUdhZzE%Yswpu2y>nyt#J1yu(=Lh z2gQ9tOI)G`6-(emh`mg?^A^YwQhI68vj)g@$G)x^ViMR0#Q!yaRmXcr_N2o1!6K4(=d*XS_wwxwkGJ z?_DWn{Vx!jB5kRTrpBo^9{_O!V@l4-CR)a;z_|{f?WylluslFm(CTT zrS!<1O1m^<8ykNz)aBP??dF5YFwI2s${wo1&>Y3bXlF(TnYh z<)$J}-QJIzd}zFUjXjORAr4qT;l_{=T+=F7%emaP@|WC>R+sO4IT?>3Wavea^;eQ*rw zt;OKFLpI1SZ5$_UQf#&RDTl}eU}Jjfl8V<76QB)Tdlj7+je(IVjF6}!yjwIRSo7DPE>FP8x3I= z&JGlPBTaueEgj7dgsD*6yaDX@i!aNVAV2`|>B`~2>(6d@H{CWwT>4mviSM+DJ}!t# zMHVLrn^x!MD-8$yvP->LyvwXN)mK}lP-Wc<2QOL{>#=&~g22%RpRn`X{ zNP!NnS#p!$pNiN4FbNgQ%ou#xN`?!lg8}7DvJ3i+T%EMt|9Evj@GQeE8Q^mv4n0H= zTCm#XpT{j-bRQAim4;VBxgWuv(R0z%gVJHuT&8E-J#mQq2&65ze61N@ZIwm=FLTP9 z@H8O+ww$cnT(}+^x0~2A&K_vJI5lNP+?P^ZHj~Cq%R@_F#ZQD(xU)sn3R_F^q0v?x zO-x=TR4tP{(#8fUojM{8;I?Yp&^;iXBaeg(|B}SeXvVCREulY5qoj_m54X#Ws%}Zs zOa7E|eHwyq4wEm@cHiYIS8e`9GSu&RDZpe+j)JD*W9PZ-{1*XC;zj zB?V1flGtiZz_Du%=*g$K&jv=0DoQH1K!nyv(I+_Pg+lvjC7Cl)l9r?aeFWiJVz)?A ziz@#!D(wC(Qm5I)R0$jgIqvC6PzYfioVxGIF==ir?Wru&ac>w0Td4L0%jIutX6X05 zgv4g1qggK@-%W`mGum>BZuBP&iAZ*He5OR1r%G8-$WkuPR`OjCx?XRZ2N9r>s>0sI zAs&)ER$5NZkb22BbPS; z8>P|Ui%6SJxI{E)1$HuUhh9?!P(l;aE=a91Li*7BRdNhXCVZcHuTHHfeMCBLYhJOh zDgG-d)toXB|A8505IfVTpypNg!mkTUjf4W8$eFXH`wVg(n&GidU0% zFT?+EanRhw=*TSZM9R1tgB`i{4ET~k&qHig=?T#ReA^i3GmQ-EH1|8Q7`L#D3YBvx zajP|A-e$WC#xI^6=?p8m4|M>B{;az1Lk)$A+V(L`7y*fkNXh#jA-@*}$@km!$8@QG zQSm$5e_r-V2cTOSS}8QV4LZ*>z7aic#Rt!g_ephE#(X^FZ&*+Ni&#HCRBG828sO2F zFfo{)cHsdo1EJ85A(u-Kib-ILQkkE#RFmdXt~PkfssO0(jjvwTe+si+7Vq3R4&>4+ zvPRl*hJY}e;1@x9`@|1;%XYd{pkpB8GDtLY1uBP}C-!P}XY|o&by=$Uuw5l{h0dIt zNPK#bj$tLu!bH^DQC%VC&^=6!dKfX;7fdEYn_m#{UIbB1lAs z#&xOW;_c%8n91AaHm)DXM^QmamGnsG>dmQ({xk8VyH+l3IcOH7Ib(B4hwEAzH#XvBliU_5O)y~OEjU}}i7y%*sx-F5zm zR&o6r3zP5^^609*T&Z-f0<2612%JBK3rl}zZ4o@i{Ek{=VyQ8}wo)nmt! z?DzyO#z}}7Pl{QobplQ{jN62s-(cz3730Bp2LHA?h~@4?dhAyn7gyyl=bnZ4a{_g| zU+3cw_HnC={BZfZqeqB6Z1NnmXvN^JF#zu#4a6ws7==+3JELr{`VU;~_+bjFKI)Ai zfUswH7L4R;0%KAWKxS=^H&i>^RX~f#>5otF12GryR1QeQ5H-s+TB5om8D$MLUW|NO zbuzfnRq(~v(943Y+OOOX4+5r*>&*L7YxSggm2B}El~6>80z5_pkal39^BQ5iD2pEn z5^@O9CFB^z9|tC~wy?vE@%Ie&NsQ97iUL#@BD_&j$-+Spv33VlxII^3@6&eLkI0a{ zLq9BY67d#~9I9F}`v3meZepzP)_YsM(8hT0Mdqm&$bx22LMdKpRGonzbU61f%d^}( zeU+W5d6tZ5XUo~{Zb_5G$cO*|8L>o|r2)6Dd+rXO72`dhVX5 z`*iu*7&0@~*A6koqc7AK9pW!qFfIC=DAEAQ>9n=ie`Ere)7bk)N7Kvz?dGr)hbTd) z;{m^kBG;h3M~-R}%U79~9;O*m0C~BfETiiGyn%(vxiZheElhMy=>G7qu^XPXfs^?| z4eCHYhMATvO?R2}AAHO4&MPS9+P#d^5m~Hv_lP|deULx<;;b8M zvQI2kS)3lQLAx856fO9;^Jy7^0KFQ}Jx0y&|Xk?39@opUl3}_vDUG z@BH|~{^j3|_vhUx&-~4|t$y?BB?}8!?%Qn>a*^sVf3LujojWJN-Rk8mzgk0JWBuu) zb$scS+NGTTO1k9t*DT5K6c9OVrX~6O^V+#PkJ#?4PM)2;sppk!&y#FTrSrcp+*zdm z!@7r0jLG6SOX#-WtFwR14Bor>)7>9|Z=Ty2pSuubA!K4Lb6$s~>F#B%{r{h%pgdF;XMeQni3XBS8*GM6+)>zUeUJz(6v{$k28F~5#FA*J=V(|IG_ zf3jo6QsMS3d09q}Mm6##`yldy5+<=eixb_;BaCD0f$m?cd(9cY}uI z7@0&Ea35{~EY!fDfe}OzanuE}f!H=$LX2ZzXka`Foc_gQAg*nf$QEGRIf1Ys6u6gy zBnwdYQ6O7{Z7C$eqPJ4ST7=0.32.0ですが、現行の実生成・GPU検証基準は0.34.2です。新しいV3.9テンプレートの実ブラウザー保存・再読込およびGPU受入は別Gateで、まだRelease/Registry公開を意味しません。 + +同一機能の旧フォルダーや旧Manager導入版を重複配置しないでください。V3.7以前のNode IDを使う保存workflowは、その対応Release/tagで開いてください。 diff --git a/LICENSE b/LICENSE new file mode 100644 index 0000000..6fd3282 --- /dev/null +++ b/LICENSE @@ -0,0 +1,21 @@ +MIT License + +Copyright (c) 2026 ukr8b3g-cmyk + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. diff --git a/MANIFEST.sha256 b/MANIFEST.sha256 new file mode 100644 index 0000000..529d05f --- /dev/null +++ b/MANIFEST.sha256 @@ -0,0 +1,477 @@ +ce319f06573b0b53709bae56d9b9ee0b3f4180eba961d7cbbac9e01e403a589e .comfyignore +3d681ded5d4c50cc20c2382e29fdc98d8a48e76d552c41ad5a195d941e66e612 .gitattributes +c3fc30c17dc92c4c772f25392f3f63385ccd5de025798d13dcea51c03f1f81e8 .github/workflows/publish_registry.yml +60392b0310924b7abcffbfa8070f0180e531a1fb1c030d85496a62b46291064e .github/workflows/publish_release.yml +3e22b441fe96bb6d099f0fcb92265f7a986c444f6c45be263c7905898aea4ac7 .github/workflows/tests.yml +a28c4c592d56c57e0967f0adce5655313b942b8cc04f77766c0db1bec96c13e0 .gitignore +fa9c99303d1a960db12ae2191cffcfbffe82e25e144eb6d16775c840fec4d573 CHANGELOG.md +6e889c42490ccb5f376694b5944c693a84e7febedf38e631d687e96bc1e76376 H3-Continuum-LLM-System-Prompt-v1.zip +144324850d4ba086938da5f4ed7b02ff0668dec554f092695328240e43fec3ec H3-Continuum-Skill-v1.zip +fe29dafce4a6c5229e7e3aef1f946afba2346015702ed8cdda9927a80b6e65f3 INSTALL_JA.txt +c44abbd789091bbead01250dbd5453e60539d2665c69f162335fdbd2936019c2 LICENSE +656e417909695c5dcabf134b6f3240c44bd02bb2e33881b5247efd345cb932c2 PACKAGE_INFO_JA.txt +57fa9ad77ab5b1d2da6ec43f7ba19118e253cb4a959b758b246755471df15591 PACKAGE_VALIDATION.txt +da45eeb9f4950777bb63974cbf1d77856b287b05e3466be63571f2ad4561f12e README.md +eed26374c7de5e9e19396836e0270fb46f37f3302dfd6f5533c8d703741cbf0f README_JA.md +9a84b03b89f5889f2ed59bcb969711bc1de367ac109b14596483283d3ffa1174 REGISTRY_MANIFEST.sha256 +0dbe8a5f71c92f3523942343b40787104c8b96781461ba3ba82934a6bdbe5834 THIRD_PARTY_NOTICES.md +02ea0951ce2e7ac2d5f3bbf02c596a9f9ea8adcdf2a113dca4c8db3cc19c36b3 __init__.py +7fd122ff08232e114a6ef32171217ea6324e980eb7257d560c466b8177a3d3ec audio_compat.py +3e7a1ec3380be2a222c8a3e9b8a6bb184d3a760478957f48186438f15235bdba branch_provenance.py +c1ed3a0342730450958fb62fd21854a876991f3a4aefb24e5383b0c73cd2312d compatibility.py +4d62508f900af9f93ec320afb900c27cc9a300d1e0c0a94714258498c5ede5fc conditioning.py +294251c40b1be2e296c5e5f627b68091d087bdb39179e3a0c3cebd2ee8a344f5 constants.py +cf255b8a635471229aebadc926acfdcacb65963adbcb9e0f9505d04eada2c6b5 continuation.py +51f1d211dfc36602a006fe5d14f4ebb08182694ad464aa9176340ff7e2bd1e41 decode_cache/__init__.py +647f55a9253976c059502dec13c7f64d7f7df1ecbe619d543c973f92f7f36776 decode_cache/identity.py +c2b44e3caed54c894556af547cd800328ed287a8cc90e0b5031f34bd6bd2224b decode_cache/nodes.py +986cc7a3dc153ff1b41c427a86756e505ea6e17d8b065df1ab15b8ff904f090e decode_cache/service.py +630ba3d2183964444c4341a8df507f7af61dac6a8e5aaca8cf387a5376d49f27 decode_cache/store.py +f9375629fa1a2c4d3538442f0c3192d7ba2f46fe9544fb2e21f0c144b53d608c docs/AB_TEST_LOG_JA.md +c93a3e52e4deabe5ef0db7f34a7605856a543089574ba17227e697e9a2c0139a docs/ARCHITECTURE.md +1d60cf5d4ef9c45885d4f90289c3b17aa70a6a2752c6b288288322a5e0568050 docs/FAILURE_SEMANTICS_BASELINE.md +c6ea788a03c5f29f32fb6a4fd62db92f108e604493d818b9729e9d7fb967526e docs/FAILURE_SEMANTICS_DELTA_R5.md +7660570840229b6fcbe62f0fbb52966d0099d64d7e6f01a739fbf9899293c8e4 docs/FUTURE_ROADMAP.md +44377cbac9131daa3d1c88ccc154c325257767f4c52f3181f61525c7eae00531 docs/ISSUE31_SAMPLING_SIGNATURES.md +b7ecd069d803cae7f86ad0ab89df144b5a5b157c168c73ee2962b47234d4ec0c docs/KJ_PREVIEW_OVERRIDE_V398.md +77c225d1e49e06d00b04acf811177c9a4317e792d4bfe92df539de99420c3563 docs/LOADER_PERSISTENCE_REPAIR.md +b30fe4640359bebfcdeeb94dbeec94cb346c8e3646509e1dcdb87d7021ba04ff docs/N2B_BRANCH_PROVENANCE_CONTRACT_V1.md +994b0b969089247921ca21306496cc1e8e6f729b587a8206b6403e0b7a11734c docs/R0_RUNTIME_CHARACTERIZATION.md +e6b8cc3250776868f473357cd97ce299fcea31ae18b879ec6e6466f47dbf704e docs/R6_NODE_FACADE_DELTA.md +bce0337cce903938b6b8df7f52c5c67e2d0ee14599c693687e4180debc606db2 docs/README_HISTORY.md +9ee92cdd3783e61e3bc71b99d5c0df7dd0a9f129c8a6b37701d7001a4f745e77 docs/README_HISTORY_JA.md +cebf8462681796df807987d8afef2c78abb0953b54b2f5dcd12d583bde0ec806 docs/REGENERATE_INTENT_V396.md +bbe74f802339b758d8874900e2617dd1a46844a58e0f872d95f34b9b40ba20ed docs/RELEASE_NOTES_V3910.md +1e3156127e570f7eeda696cf1143f42a4b3144f3ab3a32c6e2deee2ed9e8e3be docs/RELEASE_NOTES_V395.md +6de7daf79d0cf8e8bf03d01d030bc0db43df454b644c7fb277c4caa53bf0dae4 docs/RELEASE_NOTES_V396.md +8b6f43df28d660e44e6d4bdecd161157091e8f30bd2305c686918d288c9a9f1e docs/RELEASE_NOTES_V397.md +2e1885630db223593acf95c4e6ba0d6db55cf5806601fab94fe3b146307aa89f docs/RELEASE_NOTES_V398.md +49d02b7650d4a542307c9d4189dc217ba1db54c4eac959cc2e2a1ce57b501ae3 docs/RELEASE_NOTES_V399.md +91ee4adf828415e84e2f413e3a4e20dc16356f139e8ef1f6756f584cad04f3c8 docs/REVIEW_RESTORATION_REPAIR.md +93f07aa727b34802bce3d333426b8ba2af522500882a6d222d00cf577ca82e49 docs/SEAM_TEST_SOURCE_JA.md +4546ffb0cbaf270c52d1900908e46bfa2d3c0e2d0035ae5694680890e89aedd8 docs/TEMPORARY_STORAGE_JA.md +4c33d223a77c78e5c74946aa43f537031fd5aa415388d46fed975f9b07d53c3e docs/TIMELINE_VIDEO_EXPERIMENTAL.md +00f88cb57be5e79d9c881671bd13fa154d91843f4c1ae2f2451ee2a73841271d docs/TURBO_SAMPLE.md +92bee3f3c0c340356abdd06b4993fbf2e98cbe57e931f85cbfd11da0f7f87809 docs/USER_GUIDE.md +c3f0753bdae1c54044a6e9ad5770c45e85a306ac9c15cd363790f8625c865f94 docs/USER_GUIDE_JA.md +2c13aefd137db92a3dbf5576e9442d58012b1f0de1240c07d039ceeb702e7c1f docs/V2_1_4_UI_SIMPLIFICATION.md +b08022df068576c759c8bf80ac6b6519693f3c65c917f1822aeb06ea359aa5b1 docs/V2_1_5_UI_SERIALIZATION_STABILITY.md +a032738b0bc5acb136e92d8bf7694643a5f518b9c7fc0fc878d66d81ba8b1b04 docs/V2_1_6_UNIFIED_UI.md +384e3841603f7c6c8286ef147e93aaf7d056a286503a2a4fc21f397e4dec4876 docs/V34_DRIVING_AUDIO_VALIDATION.md +1840c81b4090ff3c86c0c1b0acc2cc10d3d9839e383c81f262a5c59a67504cd6 docs/V35_HIRES_FIX.md +ca9f1aecd1032f0f0b237948c7fc1e4e693440dff35f5f17fd5e6d5ca4dab7f9 docs/V38_BEGINNER_MANUAL_PLAN_AND_CAPTURE_SCRIPT_JA.md +717577efdec5e61600504b77bc2a099ef022c4fe2333e936c3310c7182d8f2a0 docs/V38_OPEN_INTEGRATION_CONTRACT.md +3496636f645b39336fd95a4f21a562810c56f8e1d92fa6eeb1e4b0786ad11b37 docs/V38_RELEASE_AND_MIGRATION.md +4d17d74107e78404b190fec9494b443249e5aff4f58670c50d54dfdfb064334b docs/V39_REFERENCE_INLINE.md +79bf8886f91eb3e8e7e8db8a8093eaa12fbb012b156e8d478de14f089f7c19f4 docs/VALIDATION_JA.md +3fec541a675c1e4e5e287cc8e498036001f34f7a62460321e16f9ba27b5e4c18 docs/VALIDATION_RESULTS.md +9a634b4ff6c9ca98285f6b250402a3d0e67d29f031e134d9bfcab54c119ed337 docs/images/comfyui-manager-h3-continuum.png +6c72d9a9e35648355c4dd4c8f060a6b7667e322f4b6c86690df9d6769ba182b7 docs/images/h3-continuum-advanced.png +b9f1dca35706a1cd963ba892d218568ea30173c6cf6a3bdc896f7a2266eb9407 docs/images/h3-continuum-detailed-report.png +de5f8fbce6bc423817acf7dc86bb0c8805a1fd962afdb2722cde7395f117eb56 docs/images/h3-continuum-image-scale.png +a384e26f3b088607c616be1a2b038365147d92da0961d8f47060fed09f795062 docs/images/h3-continuum-sampler.png +46c52e21c8f99a75c678e7abf20e381c8140860be2662bad528a6cbfc5f3295d docs/images/h3-continuum-v33-settings.png +321311da5d2c30953dc2061ca8fdd6e10b84511a15a9bc87704f000dd687cd0a docs/images/h3-continuum-v33-streamlined-ui.png +e93b6f4f8da5b2191f2d7715476eff1ab6c44fd96db689618a4bc0aa307c7b5d docs/images/v31b-auto-resume.png +c9573d4393236276d032c4503f8ed8bd19c6ef24e84eec6f64f8a4c47aabc718 docs/images/v34-driving-audio-connection.png +1b9bff0d32ab1928dd85d04700be563458484748a7bfafe4db293c4950ce6bc8 docs/images/v34-regenerate-from.png +65efc844c3f116e0a46bc35390bb46e3968df44fd44ea65a63629a3953371605 docs/images/v34-sampler-spectrum-assemble.png +ff8d901d05aa89f9b698c60b3d11534ef97982b3e22e9ede6255a85149a669c4 docs/images/v34-video-reference-separate-audio.png +06c333b3dda41526da70d56f4aa25922aafa2f48ec2d99350b0ed6b428e92bad docs/images/v34-video-reference-with-audio.png +301afedc3e0922a4bf3bcf445734b5adcdd5c35384132b24a38098c66750f499 docs/images/v34-workflow-overview.png +4e0d904d3bad89cb8c3ffe17fb645bfb6fbc87fe49fe3c75c91094b983726025 docs/images/v35-assemble-seam-node.png +a53796f7a904d675e80a2276c20d68502f03a3958529278e6ce9465baaf1c4af docs/images/v35-hires-fix-node.png +5a23f232a8b795226e8255e6e2e3248b0130f1c7ce276d50370d0333d6a62ef0 docs/images/v35-second-pass-node.png +04a34b1b55b627553d08b7b10de71295d55ea9e4ac9815b6dd8ce6d268098851 docs/images/v351-lbh-conditioning-bridge-flow.svg +b16d63242b45fe1abef0a4dddd4fc5b9306aa30c2f84bbc3ddf334a1c0832957 docs/images/v351-reference-audio-permanent.png +ed189c7f16ba1f6e48f557312066fdb5e1313b5ea7df853a99467fd0d7532ef5 docs/images/v351-video-guide-frames.png +0c1dfd5acff349b98ecdfb57ae50941f6da26a462dfd2aaa9b7c7b0a34abf304 docs/images/v352-stabilization-optimization.png +96395c3682fb0b6375dc23eacabea6ca9ee165078ff04e94458f178eef35c4ae docs/images/v36-masked-av-continuation.png +925898dfe741fdbcc9b429133cfcbfb8ff718a3b476b368a063b46d39511e2f9 docs/images/v36-sampler-node.png +edf8bcf1e9d76609a4e54ecd14a0ef6e82c9446c64355ff6ce5e25682e95a932 docs/images/v37-sampler-node.png +a8defccf63167a1bbd5db3bcba56781887f6ac5ee3f41de29683e0266d850198 docs/images/v37-tail6-second-pass.png +950daefc3ff8e71739b44a3bffe61d0566877409e8fceb1f6725908b3b169814 docs/images/v38-manual/continue-selected.png +bf14822fc2cd87ad68ef212209731befdf8962737066f9a4db16d836f554d9af docs/images/v38-manual/first-last-image-bypass.png +c90b35cd916b03d4fe65b6d835121fab7c17bb893ca44e913847f51aa42f4994 docs/images/v38-manual/full-video-progress.png +12814fe8dbde1c5efa8b33eb941dabf61a753bfe985ea7536670b5df7df2234a docs/images/v38-manual/h3-continuum-v38-civitai.png +5e4b876585fac3a2d580cdaf99b83ec3e10da90036b7923b4dffdcf0240560bd docs/images/v38-manual/h3-continuum-v38-whats-new.png +6fa8d44285dccbe84e0c24182abe4c5c87a37ae56b74d22d46131c773b5d33f4 docs/images/v38-manual/input-video-audio-bypass.png +57c024c59bdd378c3e1c97f1e0c698a2f3972b70f6336e051cc145f04ae00727 docs/images/v38-manual/live-continue-selected-steps4.png +ffef61c5ffddd3bcc014e23a66137d2d884b358d8d9d195a107c66859f9a1895 docs/images/v38-manual/live-execute-button.png +800651b2db21f69e1a7c659893208b9ce4a5d3b711a35981e11c43ca0d4c4196 docs/images/v38-manual/live-first-review.png +0ee318f1ce44d0215627a1593f696b719fa028c78cf5c01847e6b010e4479147 docs/images/v38-manual/live-regenerate-from-chunk2.png +c775220802f4c5dc43b79a047e5b921a0ba8c30d121eec201803748f66d973bd docs/images/v38-manual/live-render-history-steps4.png +c35a68cb853d23b13749144398696d575eb1ba818f771e6cd280a458b3d127f4 docs/images/v38-manual/live-retry-selected.png +05ab6bc5a54207020f708d092ade2754508b086f973b8007cc2541adc5ea90f2 docs/images/v38-manual/live-review-actions-steps4.png +0f8b4ed98acd29e83c5f1c147529bc1e306ce0d82f0602708c8e10175f9dcdc0 docs/images/v38-manual/live-review-settings.png +a0cbb625e19952fe7f3b0cdaefb11b47af6ba1b81afcd01fa1c089f6d23d398b docs/images/v38-manual/live-sequence-complete-steps4.png +84d74cda77a70edc62b869455f276c40c8587ea5785d8d97476b64b6efedcb3c docs/images/v38-manual/live-settings-complete-steps4.png +6a8421cc869fa3ec1273851cd2447e9bb38d002c4765cf0a2e08c2b38a8fd209 docs/images/v38-manual/reference-images-three-bypass.png +0be5086a48673dc79bb06027d4472022b67aacd24e1759074a39c3a0022070b9 docs/images/v38-manual/return-to-review.png +4f5a57e9a3b3d660771a9313e5231842052cd1c88f73ee21b2415858270e87ac docs/images/v38-manual/review-actions.png +5d967bc03034942d3b43061dade25a18367f81f0e1144fe03a02d8f43fc813c7 docs/images/v38-manual/review-settings.png +43d3ffe4d1f82b5f1c2876f57fd3bf9fdb99da6cfa8d02cee896862f8cf81ab3 docs/images/v38-manual/sampler-size-first-image-draft.png +c12bd55db2a25092e7c29d6d3a43d8a69452373c13e9a7559ad61c87ed4025ba docs/images/v38-manual/seed-fixed.png +04d935d4f183b14f2d25dfc7e22c2647ab47aef7826320c127dab2c1952e0de0 docs/images/v38-manual/size-source-first-image-tooltip.png +552dba25b13788d1202047b8504b8ee6447f5d99608c62741109e4f6914a46e2 docs/images/v38-manual/size-source-manual-640.png +b2b5a0736a07b9825df58fa3bc3b18f210679d16f041b1b40d984d325074c68d docs/images/v38-manual/turbo-lora-choices.png +a4705ae4b5730d2b43e29fe125a64196a031f14bcfa8702663a17aa89f683d89 docs/images/v39/v38x2-feature-summary-historical.png +ae3338f720fc22c05ab4a661c68f9c5dcb20d751a835f0e60d4738333be48e63 docs/images/v39/v39-feature-summary.png +84304d318eea6db3921d366647c6fff9ad488e8c4cfef6ca00988bfd2a6504bc docs/images/v39/v39-reference-images-node.png +b765b37502b012b462606c4d2dcef43f13950e5fd682824fb18cdb19d542c121 docs/images/v39/v39-sampler-node.png +a8c56379abe23b98c8ceed6ea5e60a02ecc3c6d9636232d7371615f4eb306e81 driving_audio.py +6771c9905d35666121c4892d2ec26202505f96d244ce814ad9a87a02878952cd examples/H3_CONTINUUM_V1_WIRING.md +c3cefcdf50df9d6609f583a008453b587c011a0fb41e290d5f1196ce62f58877 examples/H3_CONTINUUM_V2_WIRING_JA.md +be419d6c773784823af7b58a3423ef2e21d3cc3e23c40e60e274560895ca034b examples/H3_Continuum_V2_README.txt +4b94f779a98bfc555f6999113a694da07746cc3cf400bfc70ebc580592034d52 examples/H3_Continuum_V2_reference.webp +1b2212b7511fac0b63dfa6e9f006cd8644c72d08c6945ab0e4ed0e722c97a730 examples/MiniMax_H3_Continuum_V38X2.json +40199d2500174ba51c040d6983f88d74c4bfbaf19191e728818e6b969b869b0f examples/MiniMax_H3_Continuum_V39.json +705d1141842863b82abe26a8bb015d45eff3bc81ca5bdc53e58b464b743173c0 examples/MiniMax_H3_Continuum_V39_Turbo_Sample.json +68e1a724d3b99458b86b8f643dcb916183e7e8dd12d7b4b3850991feb8077694 examples/settings/recommended_settings_v1.json +5a26408c770381cdb259cb710de7dab12d7ddecca262bdbf35ba1566056050f5 examples/settings/recommended_settings_v2.json +b8275c8cecde12aee0418cea456d1e7f4b875cbb1bce32f1f3d9e06a44dd0be1 examples/v2_prompt_list_example.txt +0e59ea821081cb93165aa3f88c6949736dd225837fe1cdcf1173aa5cd54c6d29 examples/v2_prompt_timeline_example.txt +1b2212b7511fac0b63dfa6e9f006cd8644c72d08c6945ab0e4ed0e722c97a730 examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.json +4fa8680037595f55f97715f9a2f63d28f11328a87f5c33c316cc4b0ae817e668 examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.zip +1b2212b7511fac0b63dfa6e9f006cd8644c72d08c6945ab0e4ed0e722c97a730 examples/workflows/MiniMax_H3_Continuum_V38X2.json +f3d8dd0f0872afdd97fe24e58b5cb74190d03e3583a66a5d013fee41f6894cda examples/workflows/MiniMax_H3_Continuum_V38X2.zip +171f0b5596724926d1d89fd12a46434f6010f12e651d19865b9b6160a1e9fff6 examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Follow.json +e61e45f218609a7ec845d2e8762dfd1efed6c51a398061066d1e51fc93556663 examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Follow.zip +7d6754cfc75599e07046b1949c1723a56f57871ecd5abc9543fcf8b34bca8c04 examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Repeat.json +9b60c8495e3d0e32c2bcd609080342ae19d0fd4c7c0662d5fe32e7fad4a9b41c examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Repeat.zip +26bbca92b5f7cac4cbde89290b3f801b7a895df76a07493ca6f65c18d5b1154a examples/workflows/MiniMax_H3_Continuum_V38x.json +df4e0e355020d2e731932cff86a4b00932b4f193ec7ed513be0db994ec0e71bb examples/workflows/MiniMax_H3_Continuum_V38x.zip +40199d2500174ba51c040d6983f88d74c4bfbaf19191e728818e6b969b869b0f examples/workflows/MiniMax_H3_Continuum_V39.json +f69b734b5bf19fb6db78bbe4737db0bef69c2a6cef2710f28767882ef35ec847 examples/workflows/MiniMax_H3_Continuum_V39.zip +67e81fbdd4c21a8d64fb3bedb5e495698fc8df5cdd9540c8ef5476caa76119c0 examples/workflows/MiniMax_H3_Continuum_V39_Reference_Inline.json +65689fd800a60ab6939bea020d630616b03a453cf6d48022cd1361992a765abc examples/workflows/MiniMax_H3_Continuum_V39_Reference_Inline.zip +705d1141842863b82abe26a8bb015d45eff3bc81ca5bdc53e58b464b743173c0 examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.json +2096e583c6805c0f6f54f76a81e0ed1246325a16d189afba0423fa05e2c19599 examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.zip +a344da581b3c2d77558a460d2b5256e6c9a63b5008c98b52999ceda2e0b2b767 graph_contract.py +bf0556b076dbaca67c3405a7d86c333ca7fdf3d4381529b0a9edc906519943e9 guide_timeline.py +1473c384082700f422f2cf9141e0d012d87d2b43ac8a8a433edc7172e6ca7bf6 hardening.py +23bddc5a41d2e997f4fbf789c7128e32f4f197f1405b2aef1c76794f57482446 install_windows.bat +12bf572a65d57bbc9a72b459f29691c2057768f4fb7a74f9db87f4775e995bda install_windows.ps1 +139546ec1ff459945aa89b50f74b5ee69f62a0e07bbddfcae06936091a409a83 layout_adapter.py +d3e213005670e94dee33a240349a4d245cf9a3b6602099f11ed30cf5a32e2c21 media.py +7ad8ea8859b666cb2771f19795ca8c04ab7aafb5c20cff4e80ad17295429ded3 metadata.ini +fd23c7f32810788aaaf1709c4959c40ec72303aecc6f6afa5fa68d824cca97a8 minimax-h3-continuum-dance-director.zip +d7e613e2cc24110213b477bdd1d390234a41f2af3b1ed006a809066d71d658eb model_patch.py +ecb65a5dbf9e1683c5a3cadd82daaea3a90a2b776f6ebabfbe8b07aa93177422 nodes.py +6ba95639b52219fc37b4f51d52cf6879f75c7418a8d8815b3725b3f201a4f1e6 preview_override_provider_public_v1.json +d4cf474cbd1a98867075c45b53b2ab4d47752bb11611c916bac2a6831ca82c0b preview_override_provider_v1.json +3b7e18bd2ba50fd6c8fe71c58ca79a0bf752f362ed78b7357ce1998efc934b4b preview_override_signature.py +b470819f16fc179654dc1600721a63f0a5a14e777f4c004d574b499da6886ba5 public_api.py +dc56e02aa94561076bee8a836c268c0e45a8b5248f3b40108bcb740252965ec6 pyproject.toml +94fdc2d545b2d4077a4d70229fd22ba3ccac21d49d1097d2873f76d04e9a3c7a pytest.ini +caa572ea65dafc86850db99f52433d360ac9c4ab36b854292a3dccd5ba32dc76 reference.py +9539f5cb9fb695841073fa58d9246d5a7410bc372f6cd340378185004dfaf978 reference_audio.py +5cb9bfd6810d3a93a637451980a4cc5c1436e5b84cfd1c7a111cd192e876351b reference_video.py +467edb2c188d024c2392aafe7423b3f835e16b10f570da76e2f7dda8a962c1c5 requirements.txt +985e7647bd2626af5df68ca702cf6ebb38cc09f3ffe0e239d3e8ef256352a5d8 run_storage.py +8f786989af6339be8cec8f2bf602da89344cf96202d29c980af5c07175eea3d3 spectrum_provider_v029.json +8cdf37b165eb68462dd7accf4a1fd63043f02479f9c377778fb9d4b6feaddcc6 spectrum_provider_v029_py312.json +0efc8d65cb6f59042457ad843fe1b77a4bee45e59d3732f5c83fbc75a31ae77b spectrum_provider_v1.json +71801234f6d2e08b660803194653821ca648e5b9a0d0c3f138bc7ccc6fe9d8c6 spectrum_provider_v1_py312.json +448213ae7670a6e231f833ca1901ca295eb5f6bfb8983ca37703bdb3fdf15f99 spectrum_signature.py +3a359a0467725cc292ccb4c86cf0c3d30bb6a114d01f5345792b09ea3d45441d state.py +6056de6c0a5322a8a51ba5a114538dca2ef67788cd1afba10fedd483b95f6afa state_io.py +660a9df28c7d62ad383cda5514c48c7efcb3576f435af023040116d533a56563 storage_api.py +517662b104ab1507bb927a39f25b8503f092c2ba4c39bc277379ba9d9e31b334 storage_fs.py +daeda14d9f5338835661e6738e074a3897d936cd847fe3e78519efabbcdfd0ef storage_memory.py +53da3ac42f4a66e8adc4560049c7c8e7ada82599be8c61f0573c234342ca36ed storage_settings.py +43ec4023dc06c941e5d7a04db4b8a89fb58d744ceb986f6eb9bf534d4fdcb87b temporal.py +b2435d87944c3ba8669c81d729e670f1ae78a210114ed864d2acca860529787a tests/conftest.py +7f34c43ea74c43cfc4f9453848c69c8ed7546de5b78c7c7d063d643f2c483c36 tests/decode_cache/conftest.py +8d0cced8c945c4333ef78688b51a988d7e1d560e5f9e6cceec8310b17c98ad53 tests/decode_cache/helpers/cache_process_worker.py +a41a4de12e6fe5690845f9f0795fa82c4907f0afe2551034fa2a4c95617ac39c tests/decode_cache/test_adapter_nodes.py +19d47f94ed2bc56d79b4345f820de3cbb3191e47a5034363a141dd7af500408e tests/decode_cache/test_identity.py +549698d35918b611adc4e8f3aa2f0568c19291b4825cc3c593f3f63e77ec5844 tests/decode_cache/test_process_isolation.py +7e627131904dc6d6f58021fad396185c531858b62dd6febf12c3f95fd1391138 tests/decode_cache/test_store.py +38c2dbf1481d6db0afe8554a4b7c9a27c774ce8fbffcc1fcd001d35709695159 tests/fixtures/issue31_local_spectrum_signature.py.txt +cd9f3f070ab8d4486aecb28c9cf51691b02d4dad4ebf0ecb872469b180eccfa7 tests/fixtures/issue31_phase1_model_options.py.txt +db08ce77cc5c5835f2403d4d589394090fa5d25ff3b1840312b2808e1bc203b7 tests/fixtures/v38_a3c_selective_refine_saved_workflow.json +82efab13baacb9a229671eda5f4a2d24bb839b160b2bfb40591df51a5ddf80a4 tests/fixtures/v39_sampling_v5/project.json +8b8ca1b0c802d024c6915d903747d8722bea25628369997ab008ff896462b96a tests/fixtures/v39_sampling_v5/revisions/d8352287737286e5/chunks/rev-d8352287737286e5-txn-008f52b93a88461481ead25f91627af0-chunk-0002.safetensors +26e771cffc57f171f4cc4c6859876350bdf717d1e8213fc005343006a34f002f tests/fixtures/v39_sampling_v5/revisions/d8352287737286e5/chunks/rev-d8352287737286e5-txn-9bb0a584b7344342b114bd31b175225c-chunk-0001.safetensors +7bf0cf46f049e072a68e71cf8bbab10aef695f0931b16842a890c084021f5a11 tests/fixtures/v39_sampling_v5/revisions/d8352287737286e5/manifest.json +606b3985f2b644fac40c84de3e08d1453e87a988acca51ce9595516a7c60a577 tests/frontend_decode_cache_helper.mjs +868c613654467c2c739e07de5dd9b4679b36f98cd96d631d91e1d4afe5698cf5 tests/frontend_external_int_size.cjs +8d0ccf5711e1be0bf63b4e02fecb0b37cd88979894d27b89337cda4a6d1c4ec5 tests/frontend_issue30_ui.cjs +d84bc6c8f119de1c06257d1c85f135fbea74dfa075b3b9ac8d86daefcf332490 tests/frontend_reference_images_v39.cjs +c4563a9002cc17310226a27790e0ec8c1770b6525b899f398d44fe93e2b56fc6 tests/frontend_review_queue.cjs +ea47f51abe07e9b2af840272939c0f363369636af30f4c6dffb0bb696b752c8a tests/frontend_storage_settings.mjs +36f74c03a40c4b62b90eede070d976bcd1b7f0ea828983d1e61a6cb6d80a0ea8 tests/helpers/r5_process_crash_worker.py +706029af8c8d979b2b700f5bc2b96a3064fe7c728f487698d9f079abb02e9409 tests/terminal_merge_diagnostics.py +02e6286857353b76080b73928c0d1e7ec48b1ed16339231a4a450743f193315b tests/test_audit_boundary_repairs.py +041289556cc621ea1857cbb8931ae41d817f69e084b2e8a7622e184a3197b61b tests/test_compatibility.py +875ac33b5458ec5e8d845850ed495707c394436aaadb668a44f36f6a1b73764d tests/test_conditioning_modes.py +2be669b6d21ec40a8b0e172b8ab431f82681f3c97c7de0935d87ad43b0ae8b4e tests/test_context_diagnostics.py +0b06a5af3420a7157eebcca369521c8146ad546e54f85719b49562555f8f17fb tests/test_easy_load_audio.py +7a46da28616cc9fc355ae9b6a9753dadb3201932d759c086df7513ea790f3f6d tests/test_easy_load_image.py +bfcfd77e70711830d087878d78021b754cd7e1ea324d50af0caa72f041b298fc tests/test_easy_load_video.py +72c7a2ce36d8589f1de386805dba9ef1c12a3f99f9cf818e4752d5d26b39d428 tests/test_external_int_size.py +326a97499ffe51935dd3259dbab4ed71fb02955cd3ddb5a52b5c6558a7efc872 tests/test_gpu_diagnostic_runner.py +aaed0385281e1782edc96f1d084fa84bc4e7ec9e0055ec35aaf9a9f8b6e5bfce tests/test_issue13_r25_tools.py +ff2e865fb49f0c9f73f3f82eff78ff15373ce8e843514d610b73231dd6217a93 tests/test_issue13_r28_tools.py +4926706cb171e4e36e8d7ed46b592b0ed6c9c2e100072feeba497b651e52653b tests/test_issue13_r2_tools.py +d6b5b969a0c02f1b282da990db3e503e0768747cfac8a4edd32cf8677550e024 tests/test_issue30_frontend.py +6b75fc849ded1b155547b7c371f99f8928bfb180044aa0e98f46e64b1d14d805 tests/test_issue31_attention_signature.py +25683f1e109604807520b85d778947e435aefe68805ce5a1d67c30a3e51dabd8 tests/test_issue31_preview_override_signature.py +0e54b7656568ad01d06251e36b793ce912210b94c7e02f183584461f58a48de1 tests/test_issue31_spectrum_signature.py +c058f189d8e58fd48a80c20c0a857969737a8b13960eee60e5c17db2b9288bc1 tests/test_layout_adapter.py +a9da2d25e9fc9b87b249274d7bf2ab061d14f5d1e9c482b69d795ba511cdace1 tests/test_loader_store_roundtrip.py +ab1ea95e363ad0effb9459bcbe713838f12e3501515871203672ea91e7fd4c32 tests/test_m3_conditioning_display.py +a5aa420008daaa84ffb998715f52fc8c13911a6dadc43c61a860ecc455dfaee6 tests/test_model_patch.py +45e8839cd82fd955df344b3bc6efa20092b8870ea378b0c898b073665d3e47d7 tests/test_nodes_logic.py +b635d404ca67df69894c4fcf80d8782cc5699e946bfb55ae4c44fff9d930b018 tests/test_p20_file_backed_tensor_poc.py +0f18b8810b2899f87ce6d6bfeee46fcd90ba787fa09665bddd8fc128f1952c65 tests/test_p21_memory_attribution.py +2b097971258dd10f0df33e0f8cb5843ee6cd3e9ef9e7abb80b890c66baf22908 tests/test_p22_file_backed_buffer.py +d09d2bae2fac04a058ebce22fa0e761d90b267107923ab8cd01a99cc547624e6 tests/test_p23_assembly_v35.py +d2a02c120e01d360ddca17a92d5bc53b12c9f0734d06440b3fc2fad2e754e549 tests/test_p23_video_seam_patches.py +e2b000c1033f7942f21d667f20e88a17618b90ab7beaacce44ca57862d7a2f7b tests/test_p24_comfy_integration.py +c2560cbafc5962bce5f063e262cf6a173dd8df18f7b58035ec326f616617d61c tests/test_p25_gpu_acceptance_probe.py +114530f8bc4b0a6cd416581636c85d494a07e5dcca93afe9c82f4a5aaaed9dc7 tests/test_p31_prompt_conditioning_cache.py +3597bec1af56114111d778dacb05b4d0f90dbce492ee64dada704d22f566a827 tests/test_p32_api_runner.py +2e183c09d05dea5b3696d2b6a6b1811fb00c385dbb2d4b1f3dc6058af7f27971 tests/test_p32_gpu_profile_probe.py +701cb4c8b43719fac054536fc110b82dcb46939d2135d0a99313993abfc5635d tests/test_p32_profile_server_static.py +403a0ff300330ac1f97642bf0681e80b83eec18e8911c03952ddf203ce794e5a tests/test_p46a_reference_video_memory.py +2d16f6de7bcaf1f92718d3719654a746438f6218bba8d25b3632c318a1ab9c98 tests/test_packed_layout_preflight.py +cd53a068ce0df4aa4c242d5ada4609beac0d90d0b9633ada766607ea684c08fc tests/test_preview_reuse_diagnostics.py +3b9f938de680bc8abd61d1373472589df96112c50f2d46eb0f47e26c46c9f49f tests/test_preview_reuse_frontend.py +1ae0369943ba2fb13f21ad0cf56ee58621ba09bf20174379595e157a1775c7fd tests/test_public_api.py +892d3e3e022a527652dd995508c8bbb650c370caf9ee66fa3eb1f7fd288bba38 tests/test_public_workflow_integrity.py +988d4f4973410d387ef66f11619021f8548b0447759ae0dae3f559ed5c65f11f tests/test_release_metadata.py +5f6caf1814052798f20571ebb863c098bec23d8f2e091fc1f30ce1ef9c58675a tests/test_review_identity_lifecycle.py +5f7d3e01bc78fa076ecb17f8532b6d3f3137545da4028032ad6769407d80be07 tests/test_review_provenance_reroll.py +d5e255a5fdd06ea18e70f4077f855ed2f02664f91caf79b66bfbdf7df44ad07b tests/test_review_rehydration.py +ed81cc9c969f8402bf4286954e2848df5f38bd4fe16dcff2ac0b4d894a70138b tests/test_runtime_redesign_r0_characterization.py +cda09feb111efb8961e56c5f378f7925d63d0143b084c8248a9bc8be7f33ce6f tests/test_runtime_redesign_r2_validated_prefix.py +88ab786b7d45fe0b28a645da1e25213d6f971250208e7df197ecae862ab32058 tests/test_runtime_redesign_r3_pure_planner.py +5f701a84f43a04c85d6f68627defd958e7dc9a1c0023632265dd2b1d19635b87 tests/test_runtime_redesign_r4_coordinator.py +4b03e1a9140ba32b3c24e1a2567ccbe90ecbe54f1b4e1ed760bc45da5fe9fb5b tests/test_runtime_redesign_r5_transactions.py +fdaf52fbcaea3f85f4e568f4fbb217d33069d6de8b5537bd97d53fcbc3e169a0 tests/test_runtime_redesign_r6_node_facade.py +dc2f919e8540ad20272cc0bb45072e301a6d7b59e1aca22a5df9d6cf6f5cd720 tests/test_spectrum_portability.py +5af6a16a2337e3483bb0f9197fe04756decd309e4737520c5557b527d68d51bb tests/test_state.py +d4c53d2c260b635d70fa211029d608f63776e0a56c20253475fa47c58363f8a7 tests/test_state_io.py +1bbc7e5755f71d6fdfddcc6dc8337444febee87cb86773835ca6ac6c79d89357 tests/test_storage_api.py +82f049d7e51c8d6c43322ac4e0043d3fee6323c7989a2c446b86a936bd5ce790 tests/test_storage_frontend.py +a1f60f01e9d86c1f2a010752612e373e85112c7054ffc31185b9b459eb2b34a5 tests/test_temporal.py +8a512b3afdf394af043a4d6bd7c4fca0c9326e1207930bf5d96b4d0bd7205b4e tests/test_temporary_storage.py +7f518caa4d95fefd9794b7dafdb8b747963c98dad8473be73a00dff65f3c8ea4 tests/test_terminal_merge_diagnostics.py +f175e9ebf39d6fcaac578b640422676d7f7a54bfdf0f77ec44a3da6aafefd7cd tests/test_timeline_video.py +9388d33683b63dae2b51955f154d962237ddc93483437fa7c4e3c3165f17dde4 tests/test_v217_stable_facade.py +dc8b9cf0e77b4983ee4eecc60f4acde7d0e063f3b7569402feca8a56165107fd tests/test_v2_decoder.py +3cc94a6a7957ef285f58d373fbab001d926d2b43a754d60bb00bfce98675f32d tests/test_v2_prompts.py +424a4b34fb391ff3729b38a9b202e36f5bc8b9689a2a9bbf9589254c6261d9c6 tests/test_v2_registration.py +86314c190d4335586c18191dc095926909c67fe839eb29e65324acd16c44f24e tests/test_v2_seam.py +74e623120a963cb96b68651b44f47a836ea8449d30ce89d749c71ffad4766d4e tests/test_v2_seeds_motion.py +b42bcc0983147462d96c6b71cc4a38e846995dafa6eebefc3ae0d9bb36caf359 tests/test_v2_sequence_orchestration.py +0b5b20a8318862d167f135720e0489b8d4f93406d67c6646e052e35b03dcb876 tests/test_v2_session.py +b5c320279fd38d2038e39951a3d016aa9b80741523d2816e01ceda1bdecb71b8 tests/test_v2_terminal_merge.py +e22486ac62c53132c275614e4e161c78e6dce09dd0d1a7cf64f641a3b7052b99 tests/test_v2_widget_migration.py +1c84723b0c14f3169760fe1ae95d230416f6e947ada7d97ee890a2da33e9c988 tests/test_v31_run_storage.py +3af78ab85c8e19dee8d0a98bfd11a2c0e88c30d87256da1f9699c49a19a684e9 tests/test_v32_reference.py +43aef01fc587c9f9e87baf312eb4d588975aaaacab161a36bfcb638a4c11e31c tests/test_v34_driving_audio.py +aa1f36601489c360d2f1f989c10c5424bd85bd2115df2427afe696758eefd239 tests/test_v34_hybrid_reference.py +3e9a79079823bd85b0e8848c8421a71ae7aa6ad2e5d47c05c7a5cc12ef9b2101 tests/test_v351_reference_audio.py +1c317c3443435e71bc3d874d1f1ef80aca6c50263ad5e9e393923ccadd3758f7 tests/test_v351_reference_audio_ui.py +3caacf5234d96b42d708c9a742d1a7f3b4318e208358058156f97cb21b5506ed tests/test_v35_compact_ui_frontend.py +75dbdd0d9694d82fab7da23f5cc03bc118c024abc865f6c282799aa8c57427fc tests/test_v35_conditioning_bridge.py +f9e191b2262030f988b6230e2a1f32484df03568fa43ac46d2c34adb61b6dffa tests/test_v35_context_sampler.py +7380c2ea1eeb4fb212ef51ffdf6d14a33132d3ae6d4c7ed5015f28bba6b7e75b tests/test_v35_hires_fix.py +09a6f88fe4f63b12b11d28a82df1ef8eaf01c756f7bb633a516061b93ae4f73b tests/test_v35_latent_resize.py +59a07f2172014dfdc0a447775324d7cd049f20b823276dfaf9b7b3f79628f5a8 tests/test_v35_pixel_vae_resize.py +05ba609fb948665f050d58fb087569a158c040d08bebad1d728b911673b95462 tests/test_v35_refine_context.py +8ae5173eaec35adf9efcde2c464c95651c848248bdbdd6396ed2e3c3c1ac7965 tests/test_v35_refine_sampling.py +98a9006ebc4cab2b61b38904e0043fab24e088a25b28ffd38012c29991b8550c tests/test_v35_second_pass_contract.py +a11956a22f373d96f13b395971f39187d6bb12cdc153ce794b4729d258ce9725 tests/test_v35_second_pass_sampling.py +5f70cf8bd5248120ffcbe75ff3a54e23030fa6eb318ab0de6e0b08ee20e25a87 tests/test_v36_chunk_duration.py +bf69d1ac3fc30f963b6fed6cbd661f273593904b03275467427a954fbdc0a03f tests/test_v36_masked_video_prefix.py +19631b4abfdf2bd3c187ba4febbd23dd2bc6a0c5027f22c4adb8c1c4f1328926 tests/test_v36_public_sampler.py +324c213aba761e8989a78eadc0cfdcc54d098f50ed1fe84a5ea7aec26f0dc595 tests/test_v37_guide_production.py +14ce0c7f78bcba9948e9e9930c4de5a8c7e8c453fcb6172066d43f593f7175ea tests/test_v37_guide_timeline.py +70b2413ad779bdde05bb131ef19db881c48e02004287d99626467303e23f8af6 tests/test_v37_refine_schedule.py +b7da41d7654aefaa06997720f7cbbaa7ddc3827b41ed365efd5992e399aa779c tests/test_v38_adaptive_continuity_observer.py +1a24a0cb040a4effba3fcd2974df093597c1686923f0a84f0cd267305a96f7a5 tests/test_v38_audio_only_refine.py +ee4dd8f6eb8114706b2625b4471b041a0ea66198f9f825c63e913c896ed87530 tests/test_v38_branch_provenance.py +e813b5d21af2963a325862acd344b33caa124d0cd3940a940c58118cf27742e0 tests/test_v38_chunk_boundary_diagnostics.py +8cdd5e14487a59ef333cf675fe3da52eb319480e69c5a6262377533cf7d0b160 tests/test_v38_chunk_scoped_refine.py +7344eda4f112901995341ea748781920cb0ae925a665e64d9b28bbdcc810459c tests/test_v38_completed_review_ui.py +c9577e9b2d430c9435a287bf679a920020192d51f780e02b47f286e2e4b6a615 tests/test_v38_easy_mode.py +bb5a1530c890d357dc48ea74039340d746e876e8f85e06d4d9a19369f8ff5961 tests/test_v38_hover_help.py +98eb59407741dd187dc22d94b4f433f341294d05cc38106fde15b580b1465001 tests/test_v38_memory_action_policy.py +8fb637f1d61389a86ef545adf35a518a5ae44322655a38fe598fd3b3c8e0a4a9 tests/test_v38_multi_reference_audio.py +623d528a02b6712cba28c374ba3ad979bc84e2ac5318640692589a04ba6ab15f tests/test_v38_open_integration_contract.py +12337a491fa67b38b32aaeb16e0b209393960dbc46db2a8bf968b8c3902855ef tests/test_v38_packed_row_planner.py +188ec8e125b16d0bec30b63324d23501a3b89578b434f8e7b24a9d8cc43f1640 tests/test_v38_public_surface_cleanup.py +7aebd3668e04233f34c2d94e75d4708cadd64d6882cf77ec59afa0a394dcb006 tests/test_v38_public_workflow.py +14ff6561a127287d0a20887790a978612ef9d291f1d3be7d5e85b37b2c74760e tests/test_v38_ref_encode_cache.py +8b376f45261d045b584f02b4d14cd70c2d6269b9c814d4e62b5c612f6e30e36f tests/test_v38_refine_target.py +03eb5bf6d845b22334fcc800d81b419f1c8d8593fa65a1e0460144456ff2c8ba tests/test_v38_reliability.py +aa72293fe5ba6f4499e1ec95aace3b2e94813fa35cf987a767e77bbf2ae525b1 tests/test_v38_resolution_preset.py +975af1d6dd34e3403b80d3e89d8fc658baa67a34086d40cd05ac10fd92611c51 tests/test_v38_review_cache.py +72a0e3cb334049b66dd36f52c826931067c242fa63557933fd667138b38d6e8d tests/test_v38_review_control.py +f62daa235117a3041c66ebb17adedb13c0452c3a2fc0acdd51006d7aa0078a9b tests/test_v38_review_execution_cap.py +3a5e298bc9b0d46e9560e397576352117d6353dd03977ad54fe41e4881c58287 tests/test_v38_review_run_storage.py +259142da88241424afb61228b24306c88c1f99f0924e81595824ba09527e1847 tests/test_v38_review_ui.py +75697036f361a9b8ea74f69b15e408670f686c8c0e8a6c1d48fe059cc837d12a tests/test_v38_selective_second_pass_node.py +8026a241301dfc4de518649305458a538ef17a0b1567b113877a2af377738528 tests/test_v38_temporal_window_refine.py +0b2b1280a1217816984a99c0f948a52a8e2e6fdf924b5c4f2f578719b12b01ed tests/test_v38_video_audio_refine.py +68426689006a4bcdf3ae5b01589abf95c7fd3bf7f541904240412b24c0783fb0 tests/test_v38x2_decode_cache_integration.py +20aff5243985a16550d1084ce20bb35cbb91150290a3514773defc1572a6c3f1 tests/test_v39_official_workflow.py +d33a6b53e3d074d5c8a076395b050fa2b6f00e25f222c32bd78e89d3e0d2635e tests/test_v39_r0_fixed_prompt_cache.py +319258798ca198312cf684f369990924657d1af168a148ad34303af31c06a350 tests/test_v39_review_refine_scope.py +68e5ee459f87724712d450d063bb4409cb26dd112b3b62602f1b43c88a3c08fb tests/test_v39_rr_r1_fixed_identity.py +4a97449504c720b0b4846293beb857b48986b3d175215509c144ee1b52c34f23 tests/test_v39_rr_r2_reference_routing.py +32e25a0f4815f954ba3e64e2445309babc83b78a964e2a6a5287b1c7accb275e tests/test_v39_rr_r3_effective_plan.py +6ec7852560f4e0588fc0b5d34743195554475167b46606d3166fdb5fb97f9c4e tests/test_v39_rr_r4_reference_runtime.py +ddeb44c42fb3543945d908a66b0557260ce719c002afd5bae0d37794219a8841 tests/test_v39_rr_r4_sequence_integration.py +307993ea60e2265a77ddf709350f51a134fceff6e2b26dcffd893af62162262e tests/test_v39_rr_r5_reference_storage_contract.py +efe2cc9452f25dfc92f0cbe062d034c74d6121e9671a2ef0a9bb7852aa6a2652 tests/test_v39_rr_r6_public.py +49e525a86e23b26afd1008193aac510812cd2ba330d6b5b52023ab93d662ae27 tests/test_v39_stability_r1.py +b27004f82a3c551926e7e8f5bfa140af1eb2a1cf39ee5660f1d5b2495cf474ed tests/test_v39_stability_r2.py +d2e59eae0a724626b0a4e361921643dd31cb3612ae1f0d1b8b355c9d42c06174 tests/test_v39_stability_r3.py +072091309b35ff5c204c49ed478d9cdcc39d316df15a6209f687b33d13bbf790 tests/test_v39_stability_r4_r5.py +6e3f8a8a0480cc0a87a0ba7d5f79fdc1407233b155d90ff4e4b75aa57988ad49 tests/test_v39_turbo_sample.py +473abac48a57aa5f297c6e141a94de17ddcb9424a8744bae1c46e6a6a734a0ab tests/test_v3_hardening.py +9242f74663ccb04020ad2fb1199990c84433c9b4c5cf0560e93416c7b15e1913 tests/test_v3_physical_decode_groups.py +35b26c8cade7acb2eec35375d7dfa9a4a112f67a5cec6a842d61c69039de0d84 tests/test_v3_video_seam.py +0e02f7db41458fab91e6c3395bdb864d1c033923a3874ff4082d5f85b8230d98 tests/test_video_adapter.py +b07b941082a947d9e938ff399ec8b77fcd29c1a222e0e6b0dedf5d8e7482ce84 tests/test_video_reference_modes_experimental.py +f218e1a93c42467c6057c5c1d04d57cc82f6331b97ec9257e4e2dfb9c831876a timeline_video.py +0bc369ee428b8a5bc8e370a81fbab427262a1fdb7805a78c442c91698d65efba tools/__init__.py +3d5f29d1e446b94eb16bb02fea98a171c1b29248cbb4de66339030e3a7cea1c3 tools/a2_audio_only_canary_node/__init__.py +2e5a4f468c8dc70b16a083c0fac0aadc6e1e80968fcd80a3241275efbd69c669 tools/a3c_selective_refine_probe_node/__init__.py +58ed8a3e2818f5dbd2214849dc7d105fd399ec8c0b2b7c63f62102c1e18757f1 tools/audio_continuity_output_analyze.py +ea0040b510f748bc69f0c8aa4cc6aa1893e292af620ddd6e76e7847dccac1f8c tools/audio_continuity_research_runner.py +2876dc0f2760d6bffb0138ab4d6ef84da6f2db2d726cc442d7a6f732b7bc2061 tools/g4_sampling_diagnostic_node/__init__.py +9062a3b9e2d03551c8e230a34779ccd74b7e9e38aa18d19939900f0072919f82 tools/g4_sampling_diagnostic_node/node.py +0184162bf563e15b2681096732064e265c83696294e248c8a6981beb8255b568 tools/gpu_diagnostic_node/README.md +3c17d031c7f5c2a2449a0fa1e4b955b71663b043f072455350d3d25ee8e1c779 tools/gpu_diagnostic_node/__init__.py +3f9a6fb0d2f78d9b5d771b400d02c1c111a65e55ca6991da0b9dbf641a474aec tools/gpu_diagnostic_node/capture.py +f6dcddedf39be823e58dcd034e0fd583116d81f2f827860f714cb11356b9a1a4 tools/gpu_diagnostic_node/continuation_policy.py +92be5f45e1b4b3acf466d2cff24dfa119f64cdbbd5d0e22972c0a7ea9a4e4822 tools/gpu_diagnostic_node/nodes.py +5021c682783be19007ecb63775413a1272a246c128c26f314fdaff8acc2bc748 tools/gpu_diagnostic_node/soft_reanchor.py +892804c81c18249348775d233c08508c298edecabc9ff151883053f350c44906 tools/install_gpu_diagnostic_node.ps1 +68e3862e52bae7e98ee0db29f9c138db3e21007bbf4e27f1e8407c900a172df4 tools/issue13_drift_analyze.py +e922c8737c8e4a1d62299d27391c0496ecec83c0e9f26600b2039f83b8272b18 tools/issue13_drift_gate_runner.py +c315aacb3fbe6dc8a38c50cfb336760ee2fd3c293d80617360c3f3b639ae2053 tools/issue13_r25_gate_runner.py +f03916a76a1c1489fd7a23d234a04d0f26fa8afb49a409e0fea779b9fac71a11 tools/issue13_r28_analyze.py +ab4b007df62fc09a6d346dcb8c5171c8ea8b6e56e487720ef9fb91ddf19014d3 tools/issue13_r28_gate_runner.py +653d7204e48edef6caec14b57a539b54fe6fd32cb5659c0d7c43726b711a4487 tools/issue13_r2_analyze.py +55caa8df2cd57289bf39c0b9a9a43440e7c3aa2e5bb89176cb5238c15884bd8d tools/issue13_r2_gate_runner.py +2743597d289d9eb12334d1f5cdd630497e237814ddf48b17f6f68da299c01150 tools/p20_file_backed_tensor_poc.py +3bf6dfddb9a0e5276613e505465d0f01442035940c0806a3ca64daf7753a8592 tools/p24_comfy_integration_probe.py +6277e2aad67ddc9b446c9883a33e72b5194a053e5b814a9a935ce7a7770d8793 tools/p25_gpu_acceptance_probe.py +5d59c7a9c83639e1a0dc1b52ceb439586c778fbb241edd552c0f165013308bdd tools/p32_api_runner.py +46376e10e01bc7d268334d4d5be77008151b2f848fc70cc1bdab6f097e24f15a tools/p32_gpu_profile_probe.py +48503f4a81e0d1fba870a3126a7fe57e4558f0ec517b6a1369050d845e6038db tools/p32_profile_server.py +ec51ad02e69eaecfa338fe801f017ebe9d5149aa9d0f81bc0980cfd0d9374c1d tools/snapshot.ps1 +e0bddeb01d3a64e554a7db844686e6a1888d1f82232493ecf0cd931af160c73f tools/v36_fl2va_combined_gate_runner.py +48e800b407adbf96baa024ee99667d6b309f97771c0bfdf1a6e485cce7c4c3a4 tools/v36_pig3_reference_gate_runner.py +1c31c0a1d9ee9339fa91a802aed2e330ca3abdbc2bbf75268af40787fa821754 tools/v36_production_integration_gate_runner.py +a4ed2c51acf6cdb7518be9a4902ae08e6cb8c51f64dddc732cc1ed094d9ea248 tools/v36_r1_api_runner.py +79cd8ac0537c575fef40285d729e499b5ea1824706b80be7772121150fa193cf tools/v36_r1_stress_gate_analyze.py +016d8dc0d4c312632160c43e4db5a46902a91decfec01271fee46c5ba5c28452 tools/v36_r1_stress_gate_runner.py +6bbe745abe6307515ce511283d5f94321c024f1494dc1e29efd1faf6466c6fbf tools/v36_r2_audio_boundary_analyze.py +a74ebadd2f843bd11a7d3f70c0739c2ef97337040b1e54b34309d528a332df41 tools/v36_r2_cached_seam_probe.py +90662f701e7d06f32d6be7d3d1308903e497d8227ddfb74cb38103e80e6b7784 tools/v36_r2_gpu_gate_runner.py +2688126517873ebc63bc3a92151d33949e2bc3a03a56d4646176f753d3b6a9a9 tools/v38_a2_audio_only_gpu_canary.py +5d9a4a107860726548aad4e91c55807b29306f572ddffa164f3436b73e970bdc tools/v38_a3c_selective_refine_gpu_matrix.py +93e23199d1aca25402e6501abf56df17844ac6f8a0dc896a15164bc1c6410f82 tools/v38_a4_chunk_scoped_refine_gpu_matrix.py +88a8fdb844fbfbcb8e4e42f711f454cae6905a0d405a7cc5b25572aabf5d5e14 tools/v38_a6_temporal_window_refine_gpu_matrix.py +57bbb1666f09e24358ff1570f74105356c51e86d934022e6d5ea5e4c9abb0b65 tools/v38_a7a_packed_row_gpu_matrix.py +abd9ae134d78ab935cde635cae2e815e68a608add8051aaafa4e780980c23e4c tools/v38_a7b_memory_action_gpu_matrix.py +41529fa50b0a2fe0b5b318d10cfdaf63b9c670836f3b7df60ada38a432a427dd tools/v38_a8a_adaptive_continuity_gpu_matrix.py +cf9eaee6730fb15aa757b13d9a3c5775f398d845e4c7cb4be66f88f620f6a7e8 tools/v38_easy_driving_audio_gpu_gate_runner.py +c4b01f7a064490ff58d6b6d9cf599dcd29b972a5559d680073c4265e2581627e tools/v38_easy_gpu_gate_runner.py +97659b64de37568e1916699d7486aa112b94d1e2da6a1eed8fecc196b1c89459 tools/v38_g4_conditioning_determinism_gate.py +6150084050d36aad35296c3518c94f584e2de04359e2969e87fbfacf423de084 tools/v38_g4_sampling_determinism_gate.py +cc247c01a01efaf594c24c7ce36ca3357c092fc28dd742c7577093ee9d5efdef tools/v38_resolution_gpu_gate_runner.py +ad8609a45ba818fb74245b30cb0a3a221d6b345c20671092cafd714d8b07aaea tools/v38_review_phase_f_gpu_gate_runner.py +60e9bf2c7c40ef9dd71f08ec2ec8c25f7ae9b08e74c3b7be5964e1138848ae82 tools/v38_review_phase_f_recovery_runner.py +a18136a4decfeef344635bbe9bb18502231c2503b034ce1d21cc2e5e77cc5e2f tools/validate.ps1 +aab887698f4767722c833b9e55d605f9597ab0d823e32e8c5416f26a210cab1d tools/verify_runtime.py +ffd61f3f31c9b6430e1169007ee906c8f5000eba63ee44e2976ad4f5f1112671 v2/__init__.py +2003902c87bba09cfa143f5a9e4248f3a78c0fb544449a75fd40ba2cacb35655 v2/context_diagnostics.py +cde27db19b677e318cc4dfb4235be7a11429c19945f4083052a91fe117e74c82 v2/decoder.py +31012ec48a64c7467ad869c79982b7f8502c1785c33f51ee1b5727ce8b4243fb v2/diagnostics.py +2ebff4502381f675d06bba4b127b81bc3201e7579ddcd46fa5051cbdf3facd55 v2/h3_builder.py +fed4bc3cbff47bd0b16b6e99385680dc2253c745f30b7f686afafbb25f77efb1 v2/motion.py +de47a847ba27390df542086735a926d2812d77ad05b21ccc2e67633d4ffd11b0 v2/nodes.py +dc2b901789e4b024dbcbc4024c63eb8cbc564d77003ce8c35c0584851ba6dfbd v2/prompts.py +4080f1875b958b316e8e76852c4be5527a6de5c71d604e7642d3098478eb4a36 v2/sampling.py +2fd811794c4cc0f2ce050250b80484876454926b9f20fce6af3a97d728e5acdf v2/seam_guard.py +be0d8585fa57e0ccd6d02577bbc7299629e5bfe20e186747f571fccc98b67d7e v2/seam_types.py +12cfcc0c71ca087fe39a5ce2922959a62b2bb311bfe664f05c4480affce03708 v2/seeds.py +0de733c55bb61e69d48918eb5bc91318fc7cf3af383fb8937522ced9df018c47 v2/sequence.py +3ea23dbce6fcd217b6e81235855479b9b16fed7a21693cf156f9014555c4fe13 v2/session.py +13213e7709fd02bb810af3c18e4cd7d010f8cdfb121a4b99866a9f2333331089 v2/session_io.py +78c830cdf49034c6e3b2cd3da5914d5cfc5638eb63ec6dbd1bab13c50d6d4caf v3/__init__.py +ea2922567df90436dd6fa21696833a49dd0cbfeee142a5fedc3a7c303824d473 v3/adaptive_continuity.py +7815230e0a9540bc2edeaf818e940b7a1fdd817fd217ba8a4f9cb21aaf6a308e v3/assembly.py +70bec17d0dce1873b9bb4a51b09daf76833722ce89f005a22b029533a367e62b v3/assembly_v35.py +3f62fdb35e804f42d7cbf0828b4e87e69a2b99246c3f7467fe4fa698f1d803c9 v3/conditioning_bridge_nodes.py +6403c2bde3fc0b8cddef78567f663fddb48130628f4c9258fcd679c28fd0cf13 v3/driving_nodes.py +147a1d3fc5e09b323a4c3d13d30e432410389bae52b58330462375b9f307b791 v3/easy_audio_nodes.py +64b3e54b42e5156642f9957ccc374fd190d9fc7145637984e702534fb1bf80b4 v3/easy_image_nodes.py +ac859cb1a6e23b8ceae300d818c24fe1346b649a0d663017a9b3ac64aa96fca3 v3/easy_nodes.py +11e9daa45237392ac84ac1b54b3e8d5e980e77102779831824794a2a6675500b v3/easy_video_nodes.py +fe03ae0aa0dbbbe5f950bb7da34c18a078d841547992c374ea8beda5e3ad4715 v3/execution_planner.py +69bf49d0601edbe51730b9d409b1cde647feabde0bde561e416b0b472a27f4e6 v3/file_backed_buffer.py +cb53e1b104415032ec79b7e7aa1c99848e63da3eedaa251eeba102970e4172b5 v3/hires_fix_nodes.py +0858487e2b8c0060a856273ffc65d23165b2f253d183f506e27e419aea0da561 v3/latent_resize_nodes.py +5b8a54e67a608e757a505abc712effb33312636e3bb235835a485a506e80eeee v3/masked_continuation.py +91a868d2e392b018b01d0396209849c146b0d560ceb109fcdd87dea7f3d786aa v3/memory_action_policy.py +85477badd85281cfb77582fdd5b34228a9f6b68376f1b7743af2ca7d4b55aa36 v3/memory_attribution.py +4a48f74584d8ba3187e1accc31ae5acd3a2ffb9a491bb4f5ca43322db418e770 v3/node_facade.py +f2779a7453a84dec8b993cf044e69ec90da6fe86940a73181234b5000f5a8fa4 v3/nodes.py +a3411c7a295308d21623551cf4aeca5bef511eb5f063f31ac84032e3bdf4e16d v3/packed_row_planner.py +b7c4a35a5e73f528492bc6946b97b687c412f1d3f5f976c19d45c6dc20288c9a v3/pixel_vae_resize.py +7e8681bc1ca1633389e1af1ab14100338d23a047f96c59c19151acbbfae5c108 v3/plan.py +0e3dd69d6d2bec6d9491c2fb806c25f0c62629b1eb5f9e4f77ae8d27a5b19cfc v3/planning_types.py +ec71e3965f445756d54d005c0e27bea520e26dbac44e1ee3d3867bfe2da100b0 v3/ref_encode_cache.py +8b9c1cadf76e2c2425d6bf8a5783ced593a75f83a6b3f219a0f870b4c177e058 v3/reference_effective_plan.py +b64f4d3d9547ac00f8edb2155109476df5b8baea730fb94e8b5bfca6d11564e6 v3/reference_images_v39.py +4cc53a322960373949cc4e86a3102c04d3dce4dfac00f7cc62b01cf637775b62 v3/reference_plan_inspector.py +196d1b6f4a73c4fb8669e9c399ffae526bd74765f96fbf5340c780f4331e5594 v3/reference_routing.py +1366da221237853eddc9e4b724d4c28917868ea32607dd086361a84df052cfde v3/reference_runtime.py +4fb950bbdc7cdb55365fde373ee190af6cdf99f7ce4c54ef73b94f0aca4e1fda v3/reference_storage_contract.py +057921bf2b05e88c2cbc08c16e2ac483aea905b8a144fbadf4a21d6ae8252427 v3/reference_video.py +8d6cd0d7f507fd6c2d33667a6ea205e131d47571444abc94d529a98748a52517 v3/refine_context.py +519586b1b3f48c73d329f7a2aaa8a1bfc6fe31ac2ecb64f2b1fdf9c19026001a v3/refine_sampling.py +43f33f989eb70ca87b916fc33267732a2be3dd4581fc72af17c53f840b53ca49 v3/refine_schedule.py +afc149dc347b4c1502ddf32ea4229e633799cd34d2b2c6f98bf8a47910a9ef77 v3/refine_scope.py +ac373d9783d3a7bb968e8117843d246d6a090b66a819015348bff84e7747442c v3/refine_target.py +45ae91f13e8f71405d2f46045c8acc6eabbc9500c194cf5d0ba541bb085cd135 v3/refine_window.py +ae909b43ad17934c85881d3b0bea81e7b45c215eeb4d1f9f6a9ea72faf633ddf v3/reliability_v38.py +520a4ae849f83450855ce0023c582db5f91d29d8694a99aa8cf491351c837cff v3/resolution.py +3f03d8659305d2def41939205d21cdc6f35e32943274e87afe539d15b9635f9b v3/review_control.py +be6610fb9427546d9485a17015cfbdc8786715e6a0967f3e4c1dee35d8768385 v3/runtime_coordinator.py +d0a004aa3030c8adaa5ea788f9acc758fcb4a556f203f9a3f63fa4e4b5772063 v3/sampling_engine.py +97fcb293e0714c3a74b89f08e11c21dce45d25a3010948be1af9dc059d288187 v3/second_pass.py +452022107cea233aa84d5366ec4943ac5aadbd5fcf786113f1d542a641eac8f7 v3/second_pass_nodes.py +6e71e944313339f14e14e11cec682593159ca5542aa5d38095371da781d60224 v3/targeted_refine_sampling.py +5a044bdd70c546e4446e0c26af92aae88049ffa2e0e2d01b52033a83a3152d0e v3/targeted_second_pass.py +26b13614ba0a5fd8b192a106c9503b2f4b6fe7f9268a0fcb0f544979a49cab34 v3/video_adapter_nodes.py +2f07ace1462644f0bee553827b3571535de4d0c2d9ef847bd9e0ca73a4eed5cb v3/video_seam.py +4178dcfa442abdb06ce276e05bd7bcbaa6da28469ba5e57d3da3c663de4e4a9d v3/video_seam_v35.py +e124ded4bc8ddbf7b432c80d9900e40ee34214327bc6d83bd9fcf95cb55e7169 version.py +b935ad3bb729948bb3890ce967b0ecbef737a5e51ddfc47a0060fc20aa78dcda video_reference_modes.py +685df8e0fa23e3663ab31da1e4470cd19bdaa0dc16362cb606e874d9d6153c2c web/decode_cache_helper.js +fefafd4550511d80bd85ade46a0476f0cf3de13ddb94a908dff6b175f2a0e199 web/easy_bypass_toggle.js +092088096307d33ac016b82b4774ca38845a2ff2f63b97ac6da8c5aca4fea413 web/easy_load_audio.js +9db0aa94f634578a962d5db49e5c70356040a59df968e3a2e531b327a623081f web/easy_load_image.js +35090816ba7470a7255eb85d003b7e79f8a0065982dacf98a18f41c8ebff655d web/easy_load_video.js +0c68d13a1f8beb7431a32b471f8b0612c42fcc4bf72dd97611a025d158008662 web/easy_v38.js +933b5c675294a20ccec263916a1f4453d5b06a530d660ad8e56389d02d7f3023 web/project_id.js +ae59273f163a116ece418957c557c4fd8c3285e3c60814679b9361d2e6e6fa02 web/reference_audio_ui.js +46c9f75c9462bf26300332448f4431cca801f38964ca0a49bc2c50a88a7cc40f web/reference_image_ui.js +e99c74b78da1aa18be61cb053fa05e5abd106caa5eb5c814cf585786aae1e7f8 web/reference_images_v39.js +42afab2241d39c1b07ff33d8b1370d4b0f729d81e53b4c99f4bc54c6a96b8800 web/storage_settings.js +bf41bf08dd0504eb753fa5781e6c2b8c1285758ad4bb7309975bb9231d5f8e41 web/video_reference_mode_ui.js +2bcf455b4ece8b1861e327f6f9f8a79170e7c77470155aacada52e18438543dd write-h3-v39-reference-prompts.zip +3dfd0e7042070f62b9e78239c0e75d82a7df32098db10832465f3049b7f0b38f write-h3-v39-reference-prompts/SKILL.md +a9f7b67bd65a9beda622aa5fdc5f8a2ecae1b58b18690d91fb0107a16a3e384e write-h3-v39-reference-prompts/references/simple-3x5.txt +81d7b4f4da797339b46911bc96d0a3736d70df29e092715c6684bb1786908593 write-h3-v39-reference-prompts/references/structured.md diff --git a/PACKAGE_INFO_JA.txt b/PACKAGE_INFO_JA.txt new file mode 100644 index 0000000..6c410c3 --- /dev/null +++ b/PACKAGE_INFO_JA.txt @@ -0,0 +1,231 @@ +ComfyUI-H3-Continuum 3.9.9 — public KJ / Review continuation +CPU 2075 passed / 2 skipped / 2 known xfailed; upstream KJ GPU and installed local-KJ browser Continue PASS. +既存Take・Workflowは維持。更新後は再起動・ブラウザー再読込。旧unsafe/v5は自動変換しません。 +対応条件と検証範囲: docs/RELEASE_NOTES_V399.md / docs/KJ_PREVIEW_OVERRIDE_V398.md + +--- Historical package record (unchanged) --- +ComfyUI-H3-Continuum 3.9.8 / KJ Model Preview Override continuation +Full real-Core CPU gate: 1,995 passed / 3 skipped / 2 existing Inspector expected failures / zero unexpected failures and CUDA initialization attempts; the 48 dedicated KJ tests passed without skips. Normal ComfyUI GPU acceptance passed four Core-API Queues: PyTorch and Comfy Kitchen INT8 + the audited local Spectrum implementation, each with separate-Queue Chunk 1 to 2 continuation and held preview images. Both second Queues reused one chunk and generated one; original Chunk 1 bytes and contracts were unchanged. All four video/audio outputs fully decoded. See docs/KJ_PREVIEW_OVERRIDE_V398.md for exact support and limits. +Support: explicitly audited provider only; Preview Override VAE/Audio VAE disconnected, Tiny VAE none, Core NoPreviews/Auto/Latent2RGB. Restart and start a new Run. Old unsafe/v5 Takes and workflow bytes remain protected. + +--- Earlier package validation records (historical) --- + +ComfyUI-H3-Continuum 3.9.6 / Regeneration intent and automatic variations +Full CPU gate: 1,908 passed / 2 skipped / 2 known Inspector expected failures; zero unexpected failures and CUDA initialization attempts. Normal ComfyUI browser save/reload retained Chunk 2 / nonce 5 and Chunk 1 / nonce 0 after history loading. Five new-Run GPU queues passed continuation, Chunk 2 regeneration and two identical Full Run / nonce 0 submissions; earlier Chunk 1 bytes were preserved. GPU scope: PyTorch, Spectrum OFF, 512x512, 8-step Turbo; all five media outputs fully decoded. See docs/REGENERATE_INTENT_V396.md. +Update: restart ComfyUI and refresh the browser; select and save old unmarked Chunk 1 / nonce 0 again. Do not delete Runs/Takes. GitHub Release/Registry upload, service scan approval and Manager installation are separate publication checks. + +--- Previous 3.9.5 package record (historical) --- + +Previous 3.9.5 validation: 1,893 CPU passed / 2 skipped / 2 existing Inspector xfailed / failures 0; CUDA initialization 0. Public Spectrum 0.2.29 PyTorch/Kitchen Chunk 1→2 GPU continuation and full 10-second AV decode PASS. Details: docs/ISSUE31_SAMPLING_SIGNATURES.md + +ComfyUI-H3-Continuum 3.9.5 / Issue #31 保存Take継続 +パッケージmetadata: 3.9.5(2026-10-15)。公開登録は12ノード。 + +**最新版パッケージは3.9.5です。** 固定版のGitHub配布は下表を参照してください。Registryの公開状態は別に確認します。 + +| バージョン | Release | カスタムノード本体ZIP | SHA-256チェックサム | +| --- | --- | --- | --- | +| 3.9.5 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.5) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/ComfyUI-H3-Continuum-v3.9.5.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/SHA256SUMS.txt) | +| 3.9.4 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.4) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/ComfyUI-H3-Continuum-v3.9.4.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/SHA256SUMS.txt) | +| 3.9.3 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.3) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/ComfyUI-H3-Continuum-v3.9.3.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/SHA256SUMS.txt) | +| 3.9.2 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.2) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/ComfyUI-H3-Continuum-v3.9.2.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/SHA256SUMS.txt) | + +元の[3.9.2 Source ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/archive/refs/tags/v3.9.2.zip)は従来のタグから取得できます。本体ZIPと公式Workflow ZIPは別の配布物です。 + + +3.9.4はRegistry ZIPの梱包修正版です。Review復元ガイドを収録し、配布対象外のプロンプト作成資料を除外します。生成・保存動作は3.9.3と共通です。 + +設定 → H3Continuum → Storage → 一時ストレージを管理 から、サーバー側の +保存先を指定できます。FinalizeとDecode Cacheが対象で、空欄なら従来どおり。 +現在・以前の保存先、論理ファイル量、空き容量、使用中/削除可能な量を表示し、 +未使用ファイルだけを削除できます。参照中の出力・alias、所有者不明、別の生存 +プロセスや管理対象外のファイルは保護します。Run/Take・完成動画は対象外です。 + +別の物理ストレージへ書き込み先を移す機能で、総書き込み量は減りません。 +1台のディスク内の別フォルダー/別パーティションでは書き込みを逃がせません。 +HDDは遅くなる場合があります。RAMの安全余裕を検査しますが、OOM・ページング・ +容量変動や使用中のドライブ切断を完全には防げません。VRAM保存は追加しません。 + +新規ファイルから切り替え、旧出力は元の場所で保持します。Queue実行中・待機中 +の保存・削除は拒否し、マルチユーザー運用では閲覧のみです。指定先の障害時に +元のSSDへ黙って書き戻しません。既存ノードのRAM/Disk選択はそのまま使います。 + +更新はカスタムノードのみ。ComfyUI再起動とブラウザー強制再読込が必要です。 +その後の保存先設定の変更には再起動は不要です。公式Workflowのbytes、公開 +Node ID/widget順、Run Storage v3、State/Session、生成契約を維持します。 +検証結果と範囲はPACKAGE_VALIDATION.txt、詳細仕様はdocs/TEMPORARY_STORAGE_JA.md +を参照。今回GPU生成は行っていません。GitHub Release/Registry公開は別作業。 + +====================================================================== +履歴:3.9.2の記録(以下の件数・結果は当時のもの) +====================================================================== + +ComfyUI-H3-Continuum 3.9.2 / Issue #30 Frontend hotfix +パッケージmetadata: 3.9.2(2026-10-13)。公開登録は12ノード。 + +今回の修正: +- Review比較基準なしの上流snapshotを省略。 +- Reference選択時のSampler通知を1回へ整理。native値と元callbackを維持。 +- ノード全体の目標寸法と実寸が同じならsetSizeを省略。 +- 基準ありの監視、Take/Queue時の最新確認、V3.9の公開仕様・保存契約を維持。 + +検証: +Frontend修正の実Core CPUは1703成功/2スキップ/2期待失敗。 +期待失敗は独立したInspector再configure問題の2条件です。 +既存Review Queue JS 62条件と外部INT JS 22条件、Coreの登録・計画確認は成功。 +隔離Chromiumで基準なしsnapshotと通知・同寸法resizeの削減を確認しました。 +公開候補の最終試験と実ComfyUI画面の結果はPACKAGE_VALIDATION.txtを参照。 +モバイルのフリーズ解消は未確認。DOM再構築・基準ありReview走査とInspector +既知問題は残ります。今回の新規GPU生成はなく、以前のGPU結果とは区別します。 + +更新: +ブラウザーを強制再読込してください。表示するpackage番号も更新する場合は +通常の手順でComfyUIを再起動してください。引き続きV3.9ノードを使用します。 +3.9.1のRun/Takeを今回のUI修正のために移行する必要はありません。 +公式Workflow JSON/ZIPのbytesを保持。Experimental LoRA/V4.0は含みません。 +固定版はGitHub Releasesを参照し、Registry公開は別作業です。 + +====================================================================== +履歴:3.9.1安定化修正の記録(以下の件数・GPU結果は当時のもの) +====================================================================== + +ComfyUI-H3-Continuum 3.9.1 / V3.9安定化修正・V3.8X2併存 +パッケージmetadata: 3.9.1(2026-10-04) + +現行パッケージ: +- 公開登録は12ノード。引き続きSampler V3.9/Reference Images V3.9を使う。 +- V3.8X2は別Workflowとして維持。旧Run/TakeやReference配線を自動移行しない。 +- Second Passの音声のみ/複合キーフレーム、途中Reviewの条件・Reference継承を修正。 +- Sampling Contract v6/Graph Contract v4でSamplerやCFG/wrapperも照合する。 +- Review Driving Audioを元physical groupの自然時間から一度だけ切り出す。 +- Core標準resampleを優先。同一VAE内部の重みを直接変える場合はキャッシュの + 明示的クリア、またはVAE再読込が必要。 +- 公開Node ID、widget順、公式Workflow、Run Storage v3、State/Session形式を維持。 +- Experimental LoRA Plan/V4.0は含まない。 + +更新時の注意: +ComfyUIを再起動し、ブラウザーも更新する。旧v5 Takeと履歴は読めて削除されないが、 +新しいv6生成の前半には再利用できない。新規Full Video、または +「Start again from Chunk 1」で開始する。 + +検証と公開: +- 実Coreの全CPU: 1,696成功/1スキップ/失敗0。CUDA初期化0回。 +- 実CoreのRelease準備確認: 66成功。文書・Manifest確認: 9成功。 +- 公開CI: 1,676成功/21スキップ/失敗0。実Coreが必要な18件と、既存環境制約3件。 + 実Coreの1696件をCIでも実行したという意味ではない。件数を単純合算しない。 +- 対象GPUの途中Review/Chunk単独/完了経路と生成・Take再利用確認は成功。 +- ブラウザー受入と無改変公式Workflow全体の受入は別途必要。万能な画質・速度・ + アクセラレータ互換性のPASSは宣言しない。 +- GitHub Release v3.9.1は公開済み。文書訂正はmainへの後続更新として扱い、 + 公開済みtagは保持する。Registry公開は別作業で、今回は未実施。 +- Manifestは開発429件/Registry payload 144件。任意LATENT拡大の問題はIssue #27。 + +公式V3.9 Workflow: +MiniMax_H3_Continuum_V39.json/同名ZIP。32ノード/39リンク。 +画像Loader 1~6は初期Bypass、7~9は必要に応じて追加する。Sequence Promptは空欄。 +初期設定はFirst Image+Draft 0.30 MP、2×10秒、Generate Full Video。 +Queue前にモデル・画像・プロンプトを自分の素材へ設定する。Reference-onlyなら +First ImageをOFFにしてManualへ切り替える。Workflow ZIPはノード本体の +インストーラーではなく、モデル・素材も含まない。 + +現行手順: README_JA.md、docs/VALIDATION_JA.md、PACKAGE_VALIDATION.txt。 + +====================================================================== +以下は旧3.9.0および過去版の履歴。バージョン、公開状態、検証条件は当時の記録であり、 +現在の3.9.1の導入案内・受入結果ではない。 +====================================================================== + +ComfyUI-H3-Continuum 3.9.0 / V3.8X2併存 +パッケージmetadata: 3.9.0(mainソース。新しいGitHub Release/Registry公開は未実施。過去のReleaseは維持) + +現行ソース: +- 登録IDは計12件。V3.8X2の10件(旧Audio/Video Loader互換IDを含む)に、別IDのSampler V3.9とReference Images V3.9を追加 +- V3.9公式Workflowは無改変のユーザー指定MiniMax_H3_Continuum_V39.json/同名ZIP。32ノード/39リンク、画像Loader 1~6だけが新ヘルパーへ配線済み・初期Bypass。7~9は必要に応じて追加接続。V3.8X2標準ZIPにも両バージョンのJSONを同梱 +- V3.8X2公式WorkflowとSampler V3.8は別に維持。Reference配線、保存済みRun/TakeはV3.9へ自動移行しない +- 初期V3.9 WorkflowはFirst Image+Draft 0.30 MP、2×10秒、Generate Full Video(保存値はFull Run)。First Image Loaderは有効なので、Queue前に自分の画像ファイルを選ぶ。Reference-onlyならFirst ImageをOFFにしてManualへ切り替える。保存済み480×640はManual時だけ有効。Sequence Promptは空欄なので入力する +- ComfyUI依存の宣言は>=0.32.0、現行の実生成・GPU検証基準は0.34.2。新テンプレートのブラウザー保存・再読込とGPU受入は別途必要 + +以下は旧V3.8X / package 3.8.1の履歴記録です。現在の導入案内ではありません。 + +MiniMax H3の5秒ChunkをAV latent contextで継続し、長尺映像として組み立てるComfyUIカスタムノードです。 + +V3.8X / 3.8.1: +- 公開面はSampler V3.8、Finalize、Load Image、Load Audio、Load Video、Second Pass、Reference Audios helperの7ノード +- Main Samplerへfrontend-onlyのBasic/Production表示propertyを追加。Python Schema、widget順・値、backend引数は不変 +- MainのAspect選択をSize Source(First Image/Manual)へ整理。ManualではWidth/Heightを表示し、旧Aspect値は同じcanvasへ移行 +- Draft 0.30 MP、Balanced 0.60 MP、Native 768、Custom MPのResolution Resolverを維持 +- Review Each Chunk、Continue、Regenerate Current、Finish RemainingをRun Storageへ統合 +- FL2VA Terminal Merge pairを1つのatomic review unitとして維持 +- Driving AudioをReview、Smart Regenerate、Finish Remaining、restart persistenceと併用可能 +- Image/Audio/Video Loaderはnative Bypass中もEnable行を明瞭に表示し、その場でONへ復帰可能 +- 公開WorkflowはMiniMax_H3_Continuum_V38x.jsonと同じJSONを含む同名ZIP。Spectrum対応graphで、保存状態ではSpectrumとTurbo LoRAはOFF。Spectrum、rgthree、KJNodesが必要。ComfyUI-Easy-Useは不要 +- 旧V3.8.0は`v3.8.0` tagから導入可能。旧WorkflowとV3.8Xを混在させない +- 16GB GPU GateはPASS。ただしVRAM headroomは小さく、Strict Minimal 14.96 GiB、複雑なGateは約15.5~15.6 GiB +- Review途中のpartial sequenceではSecond Pass/refine_contextは正式対応外。Review完了後にSecond Passを実行 +- V3.7以前のNode IDはV3.8で非export。旧保存Workflowは対応するhistorical Release/tagを使用 +- Registry配布からtests、tools、diagnostics、Labs資産、旧Workflowを除外し、内部継承moduleは維持 + +V3.7.0: +- H3 Continuum Sampler V3.7と、1枚のStill Image Guide用optional入力を追加 +- Conditioning AdapterでFirst/Last、Continuation context、Still Guide元画像をSecond Passのtarget解像度へ再構築 +- 内部RefineSchedule v1としてExternal/Full/Tail/Partialを定義し、既存SIGMAS socketとの互換性を維持 +- 704×704 Internal Tail 6と1152×1152 Guide再ConditioningのGPU Correctness GateをPASS +- Still Image GuideはCorrectness PASS、Experimental/Production HOLD。hard anchorによる急なtrajectory変化が起きる場合あり +- V3.6 Production初期値、旧Node ID、保存済みWorkflow、Core PackedLayout/RoPE、First Pass Audio passthroughを維持 + +V3.6.1: +- Timeline/List混在PromptへH3C-P105警告を追加し、Timelineではstandalone --- 行だけを除去 +- Standard+Audio Continuity ONのFast 5/Strong 39/Autoを、実行開始時からReference Contextへ安全にfallback +- Balanced 22 Masked AV、Audio OFF Masked Video、Compatibility、Run Storage schema、旧Node/Workflow互換を維持 + +V3.6.0: +- H3 Continuum Sampler V3.6を追加 +- 標準Continuationは前ChunkのVideo/Audio latentを次Target内でNoise Mask保護するMasked AV方式 +- AdvancedのCompatibilityで従来V3.5 Reference Contextへ切替可能 +- Run Storage identityをbackend別に分離し、Terminal Mergeのlogical [2,3]はphysical group単位で再利用/再生成 +- chunk_seconds上限を30秒へ拡張(5~15秒を推奨・検証済み範囲として維持) +- V3.5.3、V3.4 Node ID、保存済みWorkflow、Second Pass、Hi-Res Fix、Assemblyを維持 + +V3.5.3: +- 公開Conditioning Bridge WorkflowのWidth/Height接続を修正 +- 公開V3.4/V3.5 Workflowの孤立link IDを清掃し、全配布Workflowのserialization整合性テストを追加 +- Hybrid First/Last+Referenceの``警告を公開番号のまま表示 +- 生成契約、Node/socket、V3.5.2最適化、V3.4互換性は不変 + +V3.5.2: +- 同一Prompt / CLIP条件の繰り返し生成でCPU conditioning結果を再利用 +- Referenceや特殊CLIP経路は安全側でcache bypass +- 長尺Video Guideの全source identityとfinite検査を維持し、必要prefixだけを保持 +- 検証条件でVideo GuideのPeak追加RSSを約71%、保持Storageを約59%削減 +- Sampling、Conditioning payload、Terminal Merge、Assembly、Seam、Run Storage、標準Seed動作は不変 + +V3.5.1: +- physical groupごとに完全なMODEL / CONDITIONINGを公開するConditioning Bridge V3.5 +- 生成音声を置換しない任意のReference Audio conditioningと常設socket +- Video loaderのIMAGE frame batchを示すVideo Guide Frames / Video Guide Size表示名 +- backend socket key、既存workflow接続、V3.4 sampling契約を維持 + +V3.5.0: +- 外部加工したH3 latentをphysical group構造のまま再SamplingするSecond Pass +- Pixel/VAE round tripを使うHi-Res Fix(Experimental) +- Latent Resize utility(高倍率の直接latent補間はMain経路非推奨) +- RAM / Disk-backed / Autoを選べるAssemble + Seam V3.5 +- Exact Duration、Seam、Terminal Merge、元音声passthroughを維持 +- V3.4ノードと保存済みワークフローの互換性を維持 + +安定範囲: +- Advanced Second Pass Bridge +- Assemble + Seam V3.5のRAM / Disk-backed / Auto + +Experimental: +- Still Image Guide V3.7(Correctness PASS、hard-anchor qualityはProduction HOLD) +- Main Hi-Res Fix V3.5 +- Main Hi-Res Fix 3x5 2x(RTX 5060 Ti 16 GiBではTerminal 77TでCUDA OOM) + +GPU受入済み: +- Internal Tail 6 704×704(外部Tail 6比Video SSIM 0.982936、decoded Audio PCM bit-exact) +- Still Guide元画像の1152×1152再Conditioning(Correctnessのみ。品質はProduction HOLD) +- Main Hi-Res Fix 1x5、576から1152 +- Hybrid FL2VA + Reference 1x5の統合Hi-Res FixとAdvanced Second Pass + +詳細: README.md および docs/VALIDATION_RESULTS.md diff --git a/PACKAGE_VALIDATION.txt b/PACKAGE_VALIDATION.txt new file mode 100644 index 0000000..f0f72d3 --- /dev/null +++ b/PACKAGE_VALIDATION.txt @@ -0,0 +1,354 @@ +ComfyUI-H3-Continuum 3.9.9 — public KJ / Review continuation +CPU 2075 passed / 2 skipped / 2 known xfailed; upstream KJ GPU and installed local-KJ browser Continue PASS. +既存Take・Workflowは維持。更新後は再起動・ブラウザー再読込。旧unsafe/v5は自動変換しません。 +対応条件と検証範囲: docs/RELEASE_NOTES_V399.md / docs/KJ_PREVIEW_OVERRIDE_V398.md + +--- Historical package record (unchanged) --- +ComfyUI-H3-Continuum 3.9.8 / KJ Model Preview Override continuation +Full real-Core CPU gate: 1,995 passed / 3 skipped / 2 existing Inspector expected failures / zero unexpected failures and CUDA initialization attempts; the 48 dedicated KJ tests passed without skips. Normal ComfyUI GPU acceptance passed four Core-API Queues: PyTorch and Comfy Kitchen INT8 + the audited local Spectrum implementation, each with separate-Queue Chunk 1 to 2 continuation and held preview images. Both second Queues reused one chunk and generated one; original Chunk 1 bytes and contracts were unchanged. All four video/audio outputs fully decoded. See docs/KJ_PREVIEW_OVERRIDE_V398.md for exact support and limits. +Support: explicitly audited provider only; Preview Override VAE/Audio VAE disconnected, Tiny VAE none, Core NoPreviews/Auto/Latent2RGB. Restart and start a new Run. Old unsafe/v5 Takes and workflow bytes remain protected. + +--- Earlier package validation records (historical) --- + +ComfyUI-H3-Continuum 3.9.7 / External-input regeneration cache +CPU: 1,948 passed / 2 skipped / 2 known Inspector expected failures; zero unexpected failures and zero CUDA initialization attempts. Targeted Core-API GPU queues verified external Chunks=2 / Nonce=0, repeated identical automatic variations 1 then 2, and 1 reused / 1 generated chunk per variation. Browser save/reload and Run-button submissions were checked separately. GPU scope: PyTorch attention, Spectrum OFF, 512x512, 8-step Turbo, two 5-second chunks. This is a targeted cache repair, not general image-quality or all-accelerator acceptance. +Existing workflow and Take files were hash-protected. No storage/schema migration. + +--- Previous 3.9.6 package record (historical) --- + +ComfyUI-H3-Continuum 3.9.6 / Regeneration intent and automatic variations +Full CPU gate: 1,908 passed / 2 skipped / 2 known Inspector expected failures; zero unexpected failures and CUDA initialization attempts. Normal ComfyUI browser save/reload retained Chunk 2 / nonce 5 and Chunk 1 / nonce 0 after history loading. Five new-Run GPU queues passed continuation, Chunk 2 regeneration and two identical Full Run / nonce 0 submissions; earlier Chunk 1 bytes were preserved. GPU scope: PyTorch, Spectrum OFF, 512x512, 8-step Turbo; all five media outputs fully decoded. See docs/REGENERATE_INTENT_V396.md. +Update: restart ComfyUI and refresh the browser; select and save old unmarked Chunk 1 / nonce 0 again. Do not delete Runs/Takes. GitHub Release/Registry upload, service scan approval and Manager installation are separate publication checks. + +--- Previous 3.9.5 package record (historical) --- + +Previous 3.9.5 validation: 1,893 CPU passed / 2 skipped / 2 existing Inspector xfailed / failures 0; CUDA initialization 0. Public Spectrum 0.2.29 PyTorch/Kitchen Chunk 1→2 GPU continuation and full 10-second AV decode PASS. Details: docs/ISSUE31_SAMPLING_SIGNATURES.md + +ComfyUI-H3-Continuum 3.9.5 / Issue #31 Attention and Spectrum identity - 2026-10-15 + +Package metadata: 3.9.5. Public registration: twelve nodes. Existing V3.9 +and separate V3.8X2 workflows remain supported without schema migration. + +Release artifacts and integrity (2026-10-13) +**Latest package: 3.9.5.** Use the fixed-version GitHub downloads below. Registry publication is verified separately. + +| Version | Release | Custom-node ZIP | SHA-256 checksum | +| --- | --- | --- | --- | +| 3.9.5 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.5) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/ComfyUI-H3-Continuum-v3.9.5.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/SHA256SUMS.txt) | +| 3.9.4 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.4) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/ComfyUI-H3-Continuum-v3.9.4.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/SHA256SUMS.txt) | +| 3.9.3 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.3) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/ComfyUI-H3-Continuum-v3.9.3.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/SHA256SUMS.txt) | +| 3.9.2 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.2) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/ComfyUI-H3-Continuum-v3.9.2.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/SHA256SUMS.txt) | + +The original [3.9.2 GitHub source archive](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/archive/refs/tags/v3.9.2.zip) remains available from its unchanged tag. This custom-node package ZIP is distinct from the official workflow ZIPs. + +3.9.4 changes package selection and version metadata only. The Registry publication workflow checks the actual ZIP file set, CRC and all declared hashes before uploading. The 3.9.3 Registry ZIP omitted docs/REVIEW_RESTORATION_REPAIR.md and included three undeclared prompt-authoring files; GitHub 3.9.3 assets are unaffected. + +Historical validation counts below retain their original version and scope. +No new GPU, long-form performance or physical SSD-write validation is claimed. + +Scope +- Server-owned scratch path for Finalize and Decode Cache. Blank inherits + existing defaults; selected storage never silently falls back to OS temp. +- Settings panel: locations, logical bytes, free capacity, explicit refresh + and conservative unused-file cleanup. No periodic filesystem polling. +- Protect live Tensor/storage aliases, cache mappings, active processes, + unknown owners, unregistered files and directory/link substitutions. +- Atomic settings/revision checks and queue/operation coordination. New + files use the selected destination; existing outputs retain their paths. +- Manual RAM and Disk-backed preflight; discoverable cgroup memory limits. + Auto's 4 GiB threshold and reserve, dtype and generation contracts remain. + +Candidate validation +- Real ComfyUI Core 0.38.0 CPU regression: 1740 passed / 2 skipped / + 2 expected failures; zero unexpected failures and CUDA initializations. + Expected failures still track the independent Inspector reconfigure bug. + Skips: Windows private-commit metric and absent historical V3.5 workflow. +- New storage regression gate: 37 cases, including real small Tensor/view + mappings, lost/changed roots, failed writes/deletes, ownerless/link + protection, detached helper leases, abrupt subprocess exit, two-process + settings conflict, HTTP queue/read-only guards, and frontend lifecycle. +- tools/validate.ps1: compileall and Core runtime verification passed using + a CPU-only Python adapter; full pytest was run separately as above. + Native PackedLayout, Fixed 3x5s planning and twelve-node registration pass. +- Actual Core 0.38.0 / frontend 1.53.6 / Chromium 151 smoke passed: real + settings navigation, invalid-path rejection, validation/save/cleanup, + ten modal reopens without polling/duplicate requests, page reload with + persisted settings, native synthetic graph serialization/reload, and + Japanese controls at a narrow viewport. No page errors or prompt submits. + The separate production-JS fixture covers 100 reopen cycles. +- All 25 tracked JSON/ZIP assets, including official workflows, retain + their pre-change bytes. Source and Registry payload hashes are verified. +- Windows-focused storage/file-backed/cache CI is configured; its hosted + result is a separate gate from these local Linux checks. The browser + gate is headless desktop Chromium, not mobile hardware or GPU generation. + +3.9.3 frontend follow-up +- Empty HTTP 404 reproduced the reported parser error and empty selector. + User startup logs identified the failed updated process binding to a port + already held by an older process. Starting the updated backend on an + available port restored status and controls in the user's environment. +- Improve HTTP/invalid-response diagnostics, initialized candidate labels, + full-width layout and manual recovery. Preserve Core URL/auth/profile + handling; no automatic retries or replay of mutating requests. +- Targeted CPU/JavaScript/HTTP regression: 59 passed / 2 existing expected + Inspector failures. Full Core 0.39.0 CPU regression: 1740 passed / + 2 skipped / 2 existing Inspector expected failures; CUDA attempts zero. +- Core 0.39.0 / frontend 1.53.10 static assets / Chromium smoke passed on + an alternate loopback host and port: initial empty-404 reproduction, + explicit HTTP 404/403/empty-200 errors, initialized full-width selector, + Refresh recovery without page reload, and actual settings save/cleanup. + No page errors, prompt submissions or automatic mutation replay. + This used Linux CPU dependencies; it is separate from Windows/GPU quality + validation. Core registration/planning and compileall also passed. +- Source/Registry hashes and unchanged workflow bytes are checked again. + +Limits and update +No GPU generation, long-form throughput, physical SSD write measurement or +mobile hardware validation was performed for this storage change. Memory +and capacity checks do not reserve resources or prevent all OOM/I/O faults. +External drives must remain connected while outputs refer to them. Unknown +legacy cache files remain protected. Multi-user UI mutations are disabled. +The existing Inspector reconfigure expected failures remain independent. +Update custom nodes, restart ComfyUI normally, then hard-refresh the browser. +Changing the saved scratch setting afterwards requires no restart. Source +publication is separate from a GitHub Release/tag or Registry publication. + +====================================================================== +Historical 3.9.2 validation (counts and scope below belong to that version) +====================================================================== + +ComfyUI-H3-Continuum 3.9.2 / Issue #30 frontend hotfix - 2026-10-13 + +Package metadata: 3.9.2. Public registration: twelve nodes. Continue using +V3.9 nodes; V3.8X2 remains a separate workflow. + +Repair scope +- Skip upstream Review settings snapshots only without a comparison baseline. +- Notify each connected Sampler once per Reference selection, preserving + synchronous native values and the original widget callback. +- Skip setSize when the complete Reference node target dimensions equal + its actual dimensions. Height-style same-value assignment remains. +- Retain Review monitoring with a baseline and fresh Take/Queue checks. +- Keep Sampling Contract v6 / Graph Contract v4, Run Storage v3, public IDs, + widget/socket order and official workflow bytes. + +Validation scopes (do not add counts from different environments) +- Frontend repair with real Core, CPU: 1703 passed / 2 skipped / 2 xfailed. + The two explicit expected failures track one independent Inspector bug: + stale verified text after 2 or 100 Sampler configure passes. +- Skips: Windows private-commit metric and absent historical V3.5 workflow. +- Existing Review Queue JavaScript: 62/62 passed; external INT: 22/22 passed. +- Offline native PackedLayout, Fixed 3x5s planning and twelve-node registration + passed. CUDA initialization attempts: zero. +- Isolated desktop Chromium, actual frontend JS/DOM with graph/API fixtures: + 600 baseline-free draw callbacks: snapshots 600 -> 0. With a baseline, + snapshots remain 600. Thirty connected checks: notifications 60 -> 30, + unchanged-dimension setSize calls 30 -> 0. Table rebuilding remains. +- Release-candidate CPU/metadata and real-ComfyUI browser results are recorded + below after final preparation. Standalone GitHub Actions is a separate gate. + +Remaining limits +Mobile freeze resolution is unconfirmed. Review snapshots with a baseline, +Reference table rebuilding and the independent Inspector display bug remain. +No new GPU generation was run for this frontend hotfix. Earlier 3.9.1 GPU +results are historical evidence, not 3.9.2 GPU validation. Full official +workflow generation and all accelerators remain outside this repair gate. + +Updating and publication +Hard-refresh the browser after updating JavaScript. Restart ComfyUI normally +if the displayed package version also needs refreshing. 3.9.1 Runs/Takes +require no migration for this UI hotfix. Users upgrading from pre-v6 code +must still follow the older v5 history instructions below. +GitHub Release/tag and Registry publication are distinct; old tags are kept. +Current publication status is listed in GitHub Releases. Registry publication +is not part of this hotfix task. + +Final 3.9.2 candidate validation - 2026-10-13 +- Real-Core CPU regression: 1703 passed / 2 skipped / 2 xfailed; no unexpected + failures; zero CUDA initialization attempts. The independent Inspector + reconfigure bug remains explicitly tracked. +- Actual ComfyUI 0.38.0 / frontend 1.53.6 / Chromium 151.0.7922.173 smoke: + physical checkbox interactions, Prompt/link/Bypass edit detection and + restoration, Chunks shrink/restore, pan/zoom, native assignment graph + serialization and app.loadGraphData reload passed. +- Before/after frontend JS responses used the actual Core/frontend/LiteGraph + with additive instrumentation. Baseline-free 600 draw callbacks: 600 -> 0 + snapshots. With a baseline: 600 -> 600. One checkbox: 2 -> 1 Sampler + notifications; same-dimension resize: 1 -> 0. +- Scope: desktop headless browser, synthetic connected image/string nodes, + model-free CPU server and graph reload. UI file-save, real stored Take + adoption, mobile hardware and full official workflow acceptance remain + separate. No POST /prompt, generation or GPU execution occurred. +- GitHub Actions status must be read from the published candidate commit; + the real-Core count above is not a claim that Core cases ran in standalone CI. + +====================================================================== +Historical 3.9.1 stability validation - 2026-10-04 +The original record below describes that package, its tests and publication. +====================================================================== + +ComfyUI-H3-Continuum 3.9.1 / V3.9 stability validation - 2026-10-04 + +Current package metadata: 3.9.1. Public registration: twelve nodes. +The V3.9 Sampler and Reference Images helper retain their public IDs and +official workflows; V3.8X2 remains a separate workflow. LoRA Plan / V4.0 is +not included. Run Storage v3 and State/Session formats are unchanged. + +Repair scope +- Preserve audio-only/mixed Second Pass keyframes; spatially adapt video only. +- Match partial/middle-chunk Review conditioning and References to the actual + output physical groups, retaining Terminal Merge groups as atomic units. +- Use Sampling Contract v6 / Graph Contract v4 for Take compatibility, + including sampler closures and MODEL CFG/wrapper settings. +- Crop Review Driving Audio once from the original physical-group time range. +- Prefer Core audio resampling. Direct weight mutation inside the same VAE + object requires explicit Reference Encode Cache clearing or VAE reload. + +Recorded validation (separate environments; do not add these counts) +- Full CPU with real ComfyUI Core: 1,696 passed / 1 skipped / 0 failed. + CUDA initialization attempts: zero. +- Release preparation with real Core: 66 passed / 0 skipped / 0 failed. +- Standalone dependency check: 48 passed / 18 skipped / 0 failed. +- Release document/manifest checks: 9 passed / 0 failed. +- GitHub Actions at release commit 94eaf70d674a58244e1cb139f021290d63fb2e6c: + 1,676 passed / 21 skipped / 0 failed. Eighteen cases require real Core and + are explicitly skipped without it; three retain existing environment or + fixture limits. This is not 1,696 real-Core cases executed in CI. +- Offline native PackedLayout, Fixed 3x5-second planning and twelve-node + registration passed. +- Targeted GPU partial-prefix, single-middle-chunk and complete-sequence + Review Second Pass checks passed at 512x512 / 24 fps, including conditioning + inheritance, source-PCM alignment and saved Take/raw preservation. +- Integrated generation and Take-reuse GPU checks passed for the repaired + runtime. These are targeted results, not universal quality/accelerator gates. + +Current manifests declare 429 development files and 144 Registry payload +files. Hash integrity does not establish semantic freshness of every document. +The official V3.9 workflow retains 32 nodes / 39 links, the original JSON +bytes and matching workflow ZIP. Models/media are not included. + +Publication and remaining gates +- GitHub Release v3.9.1 is published. Its tag remains fixed at the release + commit above; later documentation corrections are follow-up main updates. +- Registry publication is separate and has not been performed for this patch. +- Browser save/reload acceptance and full acceptance of the unchanged default + workflow remain separate gates; do not infer them from CPU or targeted GPU + results. Record the tested workflow and runtime when completing them. +- Optional latent-upscale doubled outlines remain separate in Issue #27. + +Updating: restart ComfyUI and refresh the browser. Old v5 Takes/history remain +readable and are not deleted or rewritten, but cannot supply prefixes to new +v6 generation. Start a new Full Video or Start again from Chunk 1. Continue +using the V3.9 nodes. + +See README.md, README_JA.md and docs/VALIDATION_JA.md for current instructions. + +====================================================================== +Historical V3.9.0 main-source record - 2026-09-29 +The original record below is retained as history. Its metadata, test counts, +manifest counts and publication status do not describe package 3.9.1. +====================================================================== + +ComfyUI-H3-Continuum 3.9.0 / V3.9 main-source validation — 2026-09-29 + +Status: MAIN SOURCE PUBLICATION. The ComfyUI_W worktree has 12 public node IDs: the +existing V3.8 and V3.9 Samplers plus the new V3.9 Reference Images helper. +Package metadata is 3.9.0. A GitHub main commit is not a new GitHub Release or Registry publish; earlier Releases are retained. + +The user-selected official V3.9 JSON is copied byte-for-byte and has 32 +nodes/39 links. Its matching one-JSON ZIP and gallery copy preserve that graph. +The standard V38X2 ZIP now includes both unchanged V38X2 and V3.9 JSONs for +one-download setup; the V38X2 JSON itself is unchanged. V3.9 retains the V38X2-based +model/LoRA, Decode Cache, Finalize and Save chain; Image loaders 1–6 are wired +to the new helper but bypassed by default, while 7–9 require user-added loaders. +Its Sequence Prompt is blank. Its saved size controls are First Image + Draft +0.30 MP, with the First Image loader enabled; users must select their own +image before Queue. For Reference-only, turn it off and choose Manual. +Saved 480x640 applies only to Manual mode. +Other saved controls include two 10-second chunks, Full Run mode, +Spectrum/Turbo LoRA off, and a separate Project ID. + +V3.9 connects fixed Reference Images 1–9 to one helper, then connects its one +typed output to the V3.9 Sampler. The helper's in-node table selects All chunks +or per-chunk assignments. The backend retains sparse fixed IDs, uses the +existing effective plan/Run Storage contracts, and rejects incompatible +payloads. The V3.8 Sampler and official V38X2 workflow JSON/ZIP are unchanged. +Older V3.9 prototype workflows need manual rewiring; saved Runs/Takes are not +automatically migrated. + +The latest recorded Windows CPU run is 1536 passed, 1 skipped, 0 failed. +A separate Linux/Python 3.13.5 CPU audit reported 1534 passed, 3 skipped, +0 failed; that external count was not rerun during this documentation edit. +The selected workflow JSON has 32/39 nodes/links. The development Manifest +contains 406 entries and the Registry Manifest contains 138 entries. Both +were resynchronized and hash-checked after this documentation reconciliation. +The earlier RR-R6B browser result predates this helper UI, and configured +Reference-only GPU/API evidence does not establish browser save/reload or GPU +acceptance of the current default official template. That acceptance remains HOLD. +This source publication makes no generation-code changes relative to the +validated local V3.9 candidate and does not create a tag, GitHub Release, +Registry package, or new GPU run. + +Historical V3.8X2 / package 3.8.3 validation record follows. Its test counts, +node count and workflow SHA describe that earlier preparation, not the current +V3.9 candidate: + +ComfyUI-H3-Continuum V3.8X2 / package 3.8.3 — package validation record + +Status: local commit/push preparation PASS. No commit, push, tag, GitHub Release, +or Registry publication is authorized or performed by this record. + +Public surface: exactly nine nodes +- H3 Continuum Sampler V3.8 +- H3 Continuum Reference Audios +- H3 Continuum Finalize +- H3 Continuum Load Image +- H3 Continuum Load Audio +- H3 Continuum Load Video +- H3 Continuum Second Pass +- H3 Continuum Reference Images +- Decode Cache Helper + +Official V3.8X2 workflows +- MiniMax_H3_Continuum_V38X2.json / .zip +- MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.json / .zip +- Both names contain the same graph. The current standard V38X2 ZIP also + includes V3.9; the helper-explicit ZIP remains single-JSON. The historical + V38X2 JSON SHA-256 is + 6c445d62979632bc1a801b1c28dbd639c23b51cc156ebae18837baaa757e592b. + +Reference Image 1–9 resource gate +- Tested system: NVIDIA GeForce RTX 5060 Ti 16 GB / RAM 64 GB. +- Tested run: 512 x 608, 3 x 5 seconds, Full Run, Spectrum Off, LoRA Off, + Sage Attention, Balanced 22-frame continuity. +- Nine test references were reduced to about 0.30 MP each. +- Sampling used about 15.0–15.5 GiB of the available 16,311 MiB VRAM. +- This is a high-resource configuration with little 16 GB headroom. Pre-resize + source-image copies to about 0.30 MP before loading all nine. Larger inputs, + output sizes, wrappers, or concurrent workloads may require more VRAM/RAM. + +Decode Cache Helper boundary +- It can reuse Decode only when latent contents, shape, dtype, decode metadata, + VAE identity, and a process-local cache entry all match. +- It does not accelerate Sampling. A first MISS can be slower. +- The measured 20%+ reductions are configuration-specific observations, not an + unconditional speed guarantee. + +Preparation verification +- Full CPU suite: 1,367 passed / 1 skipped / 0 failed. +- Frontend Review Queue: 46/46 PASS. +- Workflow JSON structure: 32 nodes / 38 links / zero broken saved links. +- Workflow aliases: byte-identical JSON; ZIP payload parity required. +- Development and Registry manifests: regenerated after documentation and + workflow bytes were finalized; final hash verification is part of the + commit-ready Gate. + +Registry package policy +- Include production runtime modules and only the explicitly re-included tools, + documents, and four V3.8X2 workflow files from .comfyignore. +- Exclude tests, internal AGENTS/PROJECT_STATE/WORKLOG files, development + MANIFEST.sha256, historical workflow files, caches, and unrelated ZIPs from + the Registry payload. +- MANIFEST.sha256 covers the prepared source tree except itself and the three + internal project-state files. REGISTRY_MANIFEST.sha256 covers the actual + Registry payload except itself. diff --git a/PROJECT_STATE.md b/PROJECT_STATE.md new file mode 100644 index 0000000..750a9ac --- /dev/null +++ b/PROJECT_STATE.md @@ -0,0 +1,147 @@ +# Public source status + +## 3.9.10 — Spectrum reuse and Reference assignment display + +Normalize Spectrum CRLF source auditing and add fixed Python 3.12 profiles while +preserving the existing 3.13 profiles, valid Sampling v6 identities, old Takes +and official workflows. Correct the Reference assignment display for externally +connected Chunks without changing generation count or saved selectors. + +Full local CPU/JS: 2,096 passed / 2 skipped / 2 known expected Inspector failures. +Python 3.12 component: 112 passed / 1 skip; 24 independent storage cases and +actual-JS checks passed. Reference browser checks passed. No new GPU acceptance +is claimed. Restart ComfyUI and refresh the browser; use a new Run for newly +supported configurations. See docs/RELEASE_NOTES_V3910.md. + +## 3.9.9 — KJ Preview Override / Review continuation + +Fix new-Run v6 read-only with the separately audited upstream KJ profile, and +show saved reasons when reuse is unavailable. Keep the original local profile, +Sampling v6 guards, old Takes and official workflow bytes. + +CPU: **2,075 passed / 2 skipped / 2 known expected failures**, plus overlapping +90-case JS coverage. Upstream KJ GPU and normal-ComfyUI browser Continue both +passed: **Chunk 1 preserved, Chunk 2 generated, one reused / one new**; all four +AV outputs fully decoded. Browser checks used the existing local KJ profile. +These are targeted repair checks, not general quality or all-provider acceptance. + +Restart ComfyUI and hard-refresh after updating. On Preview Override, disconnect +optional VAE/Audio VAE and set Tiny VAE `none`; retain generation VAEs. Correct +unsupported settings before starting a new Run. Old unsafe/v5 Takes remain +readable and are not automatically made reusable. No workflow rewiring or Take +deletion is required. [Full release notes](docs/RELEASE_NOTES_V399.md). + + +## V3.9 Turbo Sample on main + +Add the user-supplied sample separately: Auto, 2 x 8s, multiline image-anchored +dance prompt, fixed background/camera, Thai rap and one enabled Turbo LoRA. +Preserve the supplied 6-step/Draft settings and document their unverified pairing +with the 4-step/768p LoRA. Fresh workflow/project IDs avoid user Run identity reuse; +remove misleading auto-download metadata for the saved custom model. Standard +workflows, existing Takes and runtime code stay unchanged. No new GPU validation, +runtime deployment, release or Registry publication is included. + +Updated: 2026-10-16 (Japan time). + +## Japanese README organization (documentation only) + +The Japanese README follows the English quick-start structure. Japanese detail +and images remain in docs/USER_GUIDE_JA.md and docs/README_HISTORY_JA.md. +Downloads, model folders, provider links, history safety and limited acceptance +scope are retained. English documentation, runtime and published assets are unchanged. + +## README organization (documentation only) + +The English README is a compact entry point. Detailed operating instructions and +screenshots remain in docs/USER_GUIDE.md; release and older implementation notes +remain in docs/README_HISTORY.md. Version-specific acceptance statements retain +their historical context. Japanese README, product code, official workflows and +published v3.9.8 assets are unchanged; this is not a new runtime or release version. + +## Published v3.9.8 GPU recheck (2026-10-16) + +The exact published tag `58fbd855` was deployed and restarted in normal ComfyUI; all 464 release files and the live startup/adapter identity matched. Two new Runs / four Core-API Queues passed PyTorch/KJ and Kitchen INT8 + audited local Spectrum/KJ Chunk 1 to 2 continuation, unchanged prefix/contracts and complete AV decoding. Existing Takes/workflows stayed hash-identical. README and docs/RELEASE_NOTES_V398.md record conditions and unverified scope. The later documentation-only main commit does not replace the release tag, GitHub assets or Registry package. + +## 3.9.8 — KJ Preview Override Phase A (2026-10-16) + +A bounded adapter identifies the audited KJ Preview Override source and loaded helpers, preserves fixed settings and wrapper execution order, and excludes its reviewed transient image list. VAE/Audio VAE must be disconnected, KJ Tiny VAE must be `none`, and Core approximate-VAE preview is outside this phase. Unknown implementations remain non-reusable. Sampling v6, old Takes and official workflows are unchanged. Full real-Core CPU gate: 1,995 passed / 3 skipped / 2 existing Inspector expected failures / zero unexpected failures and CUDA initialization attempts; the 48 dedicated KJ tests passed without skips. Normal ComfyUI GPU acceptance passed four Core-API Queues: PyTorch and Comfy Kitchen INT8 + the audited local Spectrum implementation, each with separate-Queue Chunk 1 to 2 continuation and held preview images. Both second Queues reused one chunk and generated one; original Chunk 1 bytes and contracts were unchanged. All four video/audio outputs fully decoded. See docs/KJ_PREVIEW_OVERRIDE_V398.md for exact support and limits. Release and Registry state are verified separately. + +## 3.9.7 — external-input regeneration cache + +Fix automatic Full Run variations when Chunks or Variation Nonce come from external nodes. Unresolved linked inputs no longer let Core skip an explicitly requested regeneration. Earlier accepted chunks, existing Takes, public inputs and official workflows are preserved. Update the custom nodes, restart ComfyUI and refresh the browser; no workflow changes or history deletion are required. See docs/RELEASE_NOTES_V397.md for targeted validation and limits. + +## 3.9.6 — regeneration selection and automatic variations + +Package 3.9.6 preserves saved regeneration selections and routes explicit Full Run automatic variations through Run Storage on every Queue. The full CPU gate and targeted normal-runtime browser/GPU acceptance passed; scope and update steps are in docs/REGENERATE_INTENT_V396.md. Sampling v6 safety, public inputs, official workflows and stored Takes are preserved. Release assets and Registry service status are verified separately. + +## 3.9.5 — Issue #31 Attention and Spectrum identity repair + +Finite loaded-implementation signatures identify the Core PyTorch/Kitchen attention overrides and explicit Spectrum profiles, including upstream 0.2.29. Existing normal v6 contracts without the targeted override remain unchanged; original unsafe/v5 history and official workflows are preserved. Validation and compatibility limits are recorded in docs/ISSUE31_SAMPLING_SIGNATURES.md. GitHub assets and Registry service status are verified separately. + +## GitHub and Registry publication status + +**Latest package: 3.9.8.** Use the fixed-version GitHub downloads below. Registry publication is verified separately. + +| Version | Release | Custom-node ZIP | SHA-256 checksum | +| --- | --- | --- | --- | +| 3.9.8 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.8) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.8/ComfyUI-H3-Continuum-v3.9.8.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.8/SHA256SUMS.txt) | +| 3.9.7 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.7) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.7/ComfyUI-H3-Continuum-v3.9.7.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.7/SHA256SUMS.txt) | +| 3.9.6 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.6) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.6/ComfyUI-H3-Continuum-v3.9.6.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.6/SHA256SUMS.txt) | +| 3.9.5 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.5) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/ComfyUI-H3-Continuum-v3.9.5.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/SHA256SUMS.txt) | +| 3.9.4 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.4) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/ComfyUI-H3-Continuum-v3.9.4.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/SHA256SUMS.txt) | +| 3.9.3 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.3) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/ComfyUI-H3-Continuum-v3.9.3.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/SHA256SUMS.txt) | +| 3.9.2 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.2) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/ComfyUI-H3-Continuum-v3.9.2.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/SHA256SUMS.txt) | + +The original [3.9.2 GitHub source archive](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/archive/refs/tags/v3.9.2.zip) remains available from its unchanged tag. This custom-node package ZIP is distinct from the official workflow ZIPs. + +Registry 3.9.3 was uploaded; its retrieved ZIP omitted one declared document and included three undeclared prompt-authoring files. Package 3.9.4 corrects the file selection. Registry scan status is reported by the service; source publication alone does not establish scan approval. + +## 3.9.4 — Registry packaging repair (2026-10-14) + +Include the declared Review restoration guide and exclude repository-only prompt-authoring files from the Registry ZIP. The publishing workflow now builds the actual package and checks its exact file set, CRC and every Registry Manifest hash before uploading. Runtime behavior, twelve public node IDs, workflows and saved Runs/Takes are unchanged. GitHub Releases 3.9.3 and 3.9.2 remain available. + +## Package 3.9.3 — temporary storage settings + +The current source adds a server-side scratch folder setting for Finalize and Decode Cache, storage usage/free-space reporting and conservative unused-file cleanup. Manual buffer capacity checks retain the existing Auto policy. Node schemas, official workflow bytes, generation identity and saved Runs/Takes are preserved. + +Frontend follow-up improves missing/empty-response diagnostics and storage candidate initialization. The Core request address/profile handling is retained, and no automatic mutation retries are added. + +The new setting is available under Settings → H3Continuum → Storage. Validation and platform limitations are recorded in PACKAGE_VALIDATION.txt. Source publication does not imply a new GitHub Release/tag or Registry publication. + +## Previous package 3.9.2 — Issue #30 frontend hotfix + +This package includes three minimal frontend changes: skip Review snapshots without a comparison baseline, notify connected Samplers once per Reference selection, and skip unchanged node dimensions when resizing. Review baseline monitoring, synchronous Take/Queue checks and generation/storage contracts are retained; package metadata is 3.9.2. + +CPU/JavaScript regression checks passed 1,703 cases, with 2 skips and 2 explicit expected failures for the separate Inspector reconfigure bug. Offline twelve-node/runtime checks, isolated Chromium checks and the actual Core/frontend/LiteGraph desktop smoke passed without GPU generation. The live smoke covers native assignment graph serialization/reload and upstream edit/restore using synthetic nodes; UI file-save and stored-Take acceptance remain separate. DOM rebuilds remain; real ComfyUI browser save/reload, mobile freeze resolution and full workflow acceptance are still pending. No delayed update timer is added. See GitHub Releases for the published version; Registry publication is separate. + +## Previous package 3.9.1 — generation stability history + +This patch stabilizes V3.9 generation and Review without adding LoRA Plan or V4.0. Public V3.9 node IDs and official workflows remain unchanged. V3.8X2 stays available as a separate workflow; historical releases and tags remain available. + +The included repairs cover: + +- Second Pass audio-only and mixed keyframes, with video-only spatial adaptation. +- Conditioning and Reference inheritance for the actual Review output, including partial output and a single middle chunk. +- Take compatibility using Sampling Contract v6 and Graph Contract v4, including sampler closures and MODEL CFG/wrapper settings. +- Review Driving Audio cropped once from the physical group's natural frame range, sharing the PCM origin for Exact ON/OFF. +- Core-standard audio resampling with the legacy fallback when the API is unavailable. +- The documented Reference Encode Cache limitation for direct weight mutation inside the same VAE object. + +## Compatibility + +Original v5 Takes and history remain readable and are not deleted or converted. They cannot supply prefixes for v6 generation. Start a new Full Video or use Start again from Chunk 1 when moving an older Run to the repaired runtime. + +Run Storage v3 and State/Session formats are retained. Unknown upstream settings can still generate; when their identity cannot be observed safely, automatic Take reuse is disabled. + +## Validation and remaining limits + +- Full CPU regression: 1,696 passed, 1 skipped, no failures; CUDA initialization attempts: zero. +- Offline verification: native PackedLayout and Fixed 3x5-second planning passed; twelve public nodes registered. +- Targeted GPU checks passed partial-prefix Review, single-middle-chunk Review and complete three-group reuse at 512x512 / 24 fps. Corresponding saved FLAC matched source PCM, and original Take/raw files were retained. +- Integrated generation and Take-reuse GPU checks passed for the repaired runtime. +- Browser save/reload acceptance and full acceptance of the unchanged default workflow remain pending. These targeted results do not establish every accelerator or quality combination. +- Optional latent-upscale doubled outlines are tracked separately in [Issue #27](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/27). +- Versioned source downloads are listed in GitHub Releases. Registry publication is separate, and Release publication does not complete the pending acceptance gates. + +See README.md, README_JA.md and CHANGELOG.md for user instructions and repair details. Current package records are in PACKAGE_VALIDATION.txt and PACKAGE_INFO_JA.txt; docs/VALIDATION_JA.md separates the current acceptance procedure from historical V2 instructions. Follow-up documentation corrections on main retain the published v3.9.1 tag. diff --git a/README.md b/README.md new file mode 100644 index 0000000..7d03b76 --- /dev/null +++ b/README.md @@ -0,0 +1,306 @@ +# ComfyUI-H3-Continuum 3.9.10 — V3.9 + +Generate long-form MiniMax H3 video and audio in chunks, review each result, +retry a section, and continue from compatible saved Takes. +**V3.9 is the current workflow; V3.8X2 remains available separately.** + +[Download v3.9.10](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.10/ComfyUI-H3-Continuum-v3.9.10.zip) · +[Release notes](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.10) · +[Checksums](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.10/SHA256SUMS.txt) · +[日本語](README_JA.md) + +## 3.9.10 — Spectrum reuse and Reference assignments + +Fix false Sampling v6 read-only caused by Spectrum source CRLF line endings, +and add separate audited Python 3.12 profiles for the two supported Spectrum +implementations. Existing Python 3.13 profiles and valid Take identities stay exact. +Unknown code or runtime state remains non-reusable. + +With external Chunks inputs, the V3.9 Reference Images table shows an assignment +range of 1–16 instead of a stale widget count. This does not change the number +of generated chunks or expand saved selections automatically. + +CPU/JS: **2,096 passed / 2 skipped / 2 known expected failures**. Separate +Python 3.12 component, storage and actual-JS checks passed. Reference table +browser acceptance passed. **No new GPU acceptance is claimed for 3.9.10.** + +Restart ComfyUI and refresh the browser after updating. Keep existing workflows +and Takes. Start a new Run for a newly supported configuration; old unsafe/v5 +Takes remain readable and are not automatically made reusable. +[Release notes and limits](docs/RELEASE_NOTES_V3910.md). + +## 3.9.9 — KJ Preview Override / Review continuation + +Fix new-Run v6 read-only with the separately audited upstream KJ profile, and +show saved reasons when reuse is unavailable. Keep the original local profile, +Sampling v6 guards, old Takes and official workflow bytes. + +CPU: **2,075 passed / 2 skipped / 2 known expected failures**, plus overlapping +90-case JS coverage. Upstream KJ GPU and normal-ComfyUI browser Continue both +passed: **Chunk 1 preserved, Chunk 2 generated, one reused / one new**; all four +AV outputs fully decoded. Browser checks used the existing local KJ profile. +These are targeted repair checks, not general quality or all-provider acceptance. + +Restart ComfyUI and hard-refresh after updating. On Preview Override, disconnect +optional VAE/Audio VAE and set Tiny VAE `none`; retain generation VAEs. Correct +unsupported settings before starting a new Run. Old unsafe/v5 Takes remain +readable and are not automatically made reusable. No workflow rewiring or Take +deletion is required. [Full release notes](docs/RELEASE_NOTES_V399.md). + +## Getting started video + +[Watch: H3 Continuum V3.9 — Chunks, Prompts & Reference Images](https://youtu.be/AKJxBXaiG6Q) — **12:26**, with English narration, subtitles and chapters. + +The video covers First Image setup, chunk settings, Review/Take retries, +Timeline prompts and Reference Images. For screenshots and every control, +open the [illustrated user guide](docs/USER_GUIDE.md). + +## Install or update + +Use one Continuum installation in `ComfyUI/custom_nodes`. +Models and third-party custom nodes are not bundled. + +### Fixed-version ZIP + +1. Download the **custom-node ZIP** above and optionally verify its checksum. +2. Extract its package folder into `ComfyUI/custom_nodes`. + When updating, replace the existing package; do not leave duplicate copies. +3. Restart ComfyUI and hard-refresh the browser (**Ctrl+F5**). +4. Open a V3.9 workflow below and select your installed models and input files. + +Keep your saved workflows, Run/Take folders and settings when updating. +**Workflow ZIPs contain graphs, not the custom-node installer.** + +### Git or ComfyUI-Manager + +For a new Git installation: + +```bash +cd ComfyUI/custom_nodes +git clone https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum.git +``` + +For an existing Git installation: + +```bash +cd ComfyUI/custom_nodes/ComfyUI-H3-Continuum +git pull --ff-only origin main +``` + +If Manager owns your installation, update that copy through Manager. +Restart ComfyUI and refresh the browser after updating. +Registry review, Manager's selected version and GitHub Releases are separate; +the fixed-version links above identify the exact v3.9.9 package. + +The package declares Python **3.10+** and ComfyUI **0.32.0+**. +The minimum declaration is not a GPU acceptance result for every Core version. + +## Required model files and extensions + +Download the weights separately from [Comfy-Org / MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3). +The table follows the **saved V3.9 template selections**; folders are relative to `ComfyUI/`. + +| Component | File / download | Folder | +| --- | --- | --- | +| H3 diffusion model | [minimax_h3_fl2va_pruned_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors) | `models/diffusion_models/` | +| Text encoder | [qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors) | `models/text_encoders/` | +| Video VAE | [minimax_h3_video_vae_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_int8_convrot.safetensors) | `models/vae/` | +| Audio VAE | [minimax_h3_audio_vae_fp32.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors) | `models/vae/` | + +Select the downloaded files in the loaders; set the text encoder's type to `minimax`. +Reselect paths if the saved `minimax/` subfolder differs from your installation. +The Video VAE loader selects INT8, although its older embedded download metadata names FP16; +choose the VAE you actually installed. [Other model variants](https://huggingface.co/Comfy-Org/MiniMax-H3/tree/main/diffusion_models) must match the generation task and loader support. +Quantized variants require compatible Core/GPU support; these selections are not a universal hardware guarantee. +The retained V3.8X2 graph instead saves DasiwaMinimaxH3_dasiwaHybridV2_int8.safetensors and +[FP16 Video VAE](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors). +That custom model is not bundled or supplied by the table above; reselect a compatible H3 model you have installed. + +The complete supplied graph also needs these external node packages: + +- [Spectrum](https://github.com/xmarre/ComfyUI-Spectrum-MiniMax-H3) — optional acceleration; saved disabled. Its presence does not imply every version is reuse-compatible; see the compatibility table below. +- [KJNodes](https://github.com/kijai/ComfyUI-KJNodes) — supplies the template's H3 SageAttention patch. Follow its attention dependency requirements, or replace/bypass that patch if unavailable. +- [rgthree-comfy](https://github.com/rgthree/rgthree-comfy) — supplies `Power Lora Loader (rgthree)`. + +**Optional Turbo:** download one task-matched [LightX2V Turbo LoRA](https://huggingface.co/lightx2v/Minimax-h3-Turbo) into `models/loras/`; the standard workflows start with all entries disabled. The separate Turbo Sample enables one LoRA. +Follow that LoRA's sampler/step instructions; [listed files and setup](docs/USER_GUIDE.md#turbo-lora-and-spectrum-in-supplied-workflows) explain the choices. Models and extensions are not installed by Continuum. + +## Choose a workflow + +| Workflow | Download | Use | +| --- | --- | --- | +| **V3.9** | [JSON](examples/workflows/MiniMax_H3_Continuum_V39.json) / [ZIP](examples/workflows/MiniMax_H3_Continuum_V39.zip) | Current Sampler and fixed Reference Image slots | +| **V3.9 Turbo Sample** | [JSON](examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.json) / [ZIP](examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.zip) | Auto, 2 x 8s dance and Thai rap; [setup and limits](docs/TURBO_SAMPLE.md) | +| **V3.8X2** | [JSON](examples/workflows/MiniMax_H3_Continuum_V38X2.json) / [combined ZIP](examples/workflows/MiniMax_H3_Continuum_V38X2.zip) | Retained workflow; combined ZIP also contains V3.9 | + +They are also available in **Templates → ComfyUI-H3-Continuum**. +V3.8X2 graphs and their Reference wiring are not automatically converted to V3.9. +Keep the original graph and open the dedicated V3.9 graph for new work. + +The templates reference external Spectrum, rgthree and KJNodes nodes. +Install, replace or remove missing optional dependencies before Queue. +In the standard workflows, Spectrum and Turbo LoRA entries are saved disabled; ComfyUI-Easy-Use is not required. + +In the V3.9 template, select your own **First Image** file before Queue. +Reference loaders 1–6 start bypassed and all per-chunk assignments start off. +Slots 7–9 require you to add and connect loaders. +See [template setup and screenshots](docs/USER_GUIDE.md#start-here-v39-and-the-retained-v38x2-workflow). + +## First generation + +Start with a small two-chunk test before a long sequence. + +1. Select your H3 model, text encoder, sampler and appropriate VAE files. +2. Write `Sequence Prompt`; use `Prompt Format = Fixed` for one shared prompt. +3. For image-to-video, use `Size Source = First Image` and choose a resolution. + The template starts at **Draft — 0.30 MP**. For exact Width/Height, use `Manual`. +4. Set `Chunks` to **2**, `Seconds per Chunk` to **5**, and check `Total Length` = **10 seconds**. +5. Keep `Base Seed` unchanged and `Control After Generate` set to `fixed`. +6. Set `Run` to `Review Each Chunk` and `Progress` to `On — Resume and Takes available`. +7. Leave `Regenerate From = Auto`, read `Ready to Queue`, then press ComfyUI's top-right **Run** button. + +`Chunks` is the final target count, not the number to add on the next Queue. +Review normally generates one review unit at a time: the first output in this +example is **5 seconds**, even though the target is 10 seconds. +With Last Image terminal merging, the final two logical chunks may form one unit. + +For uninterrupted generation, select `Generate Full Video` instead. +Keep Progress on if you want saved Takes, compatible resume and later regeneration. + +## Review, continue and retry + +Selecting a Review action does not start sampling. Select it, then press +ComfyUI's **Run** button (`Queue`) to execute it. + +| Action | Next Queue | +| --- | --- | +| `Use it and continue` | Keep accepted chunks and generate the next review unit | +| `Try this chunk again` | Keep earlier chunks and create another Take of the reviewed unit | +| `Use it and finish the rest` | Keep accepted chunks and generate all remaining units | +| `Back to Settings` / `Return to Review` | Navigate between settings and review; neither queues work nor deletes Takes | +| `Use This Take` | Select a stored result without new sampling; decoding/saving may still run | +| `Continue From Here` | Branch from the selected Take and generate the next review unit | + +For the **2 × 5-second** example, continue after Chunk 1 to reuse it and generate +Chunk 2; the completed result is **10 seconds**. +At completion, retrying the current unit creates another Chunk 2 Take. + +Keep the same model/LoRAs, seed, prompts, media, dimensions, sampling schedule, +continuation settings and Run identity while continuing. +Keeping only the seed fixed is insufficient to guarantee reuse. +Read the status/reuse counts before treating a result as a successful continuation. + +To extend a completed sequence, keep those conditions and `Regenerate From = Auto`, +increase the final `Chunks` target, and Queue again. +[Illustrated Review, Take and extension steps](docs/USER_GUIDE.md#review-continue-and-revise-a-chunked-video). + +## Regenerate from a chosen chunk + +1. Open settings and choose **Regenerate From = Chunk N**. +2. Keep Progress on and earlier-chunk generation inputs unchanged. +3. Choose `Generate Full Video` to rebuild N through the target in one Queue, + or `Review Each Chunk` to pause after the next review unit. +4. Set **Variation Nonce** and press ComfyUI's **Run** button. +5. Restore `Regenerate From = Auto` before ordinary continuation or Take actions. + +Nonce **0** requests another automatic variation after a completed regeneration; +a compatible interrupted regeneration resumes its existing variation. +A positive Nonce fixes that variation: repeating it with identical inputs does +not request a new result. Review retry manages its own variation separately. + +Earlier compatible chunks and previous Takes are retained. +Later chunks follow the changed continuation, so Full Run from Chunk 2 may +regenerate more than Chunk 2 when the target contains additional chunks. +Terminal-merged pairs are atomic; arbitrary frames inside a chunk cannot be edited. +[Regeneration details and compatibility](docs/REGENERATE_INTENT_V396.md). + +## Prompts and Reference Images + +- **Fixed:** one shared prompt; **List:** per-chunk prompts; **Timeline:** time-based instructions. +- V3.9 Reference Images uses fixed **Image 1–9** slots and one Sampler connection. + `All chunks` shares connected images; `Per chunk` lets you tick assignments. +- `@R1` always means Image 1, even if other slots are empty. Use tags only for + active References; resolved picture numbers can differ by physical group. +- First/Last Image, Reference Images, Reference Audio and Driving Audio have + different roles. Driving Audio preserves a source track; Reference Audio conditions generation. +- For a Reference-only test, turn First/Last Image off and choose Manual size. + A terminal-merged final pair needs the same effective Reference set in both chunks. +- Pre-resize large source images: conditioning resize does not eliminate + source-loading RAM use. A subject may persist through video continuity after its assignment ends. + +[Reference guide](docs/V39_REFERENCE_INLINE.md) · +[Prompt examples](docs/README_HISTORY.md#prompt-formats) · +[V3.9 prompt skill](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/tree/main/write-h3-v39-reference-prompts) · +[General prompt resources](docs/USER_GUIDE.md#prompt-and-skill-downloads) + +## Storage, decoding and Second Pass + +Progress saves Run/Take data for compatible reuse. The optional **Decode Cache +Helper** caches decoded results; it is separate from saved sampling history. +Finalize assembles the output; downstream Core nodes create and save the video. + +Set scratch storage under **Settings → H3Continuum → Storage → Manage temporary storage**. +The folder is on the ComfyUI server and applies to Finalize and Decode Cache. +It does not move saved Runs/Takes or finished videos. +Change or clean it only with empty running/pending queues. +Moving folders on one physical drive does not reduce that drive's writes. +[Storage modes, cleanup and limits](docs/TEMPORARY_STORAGE_JA.md). + +**Second Pass** is an optional bridge for external refinement/upscaling. +It is separate from chunk continuation and ordinary Review retries. +The V3.8X2 standard workflow does not require Hi-Res Fix. +V3.8 Second Pass does not support partial Review sequences; V3.9 can inherit +validated conditioning for its actual output groups. +[Integration contract](docs/V38_OPEN_INTEGRATION_CONTRACT.md). + +## Compatibility and v3.9.9 scope + +| Area | Supported boundary | +| --- | --- | +| Saved history | Existing Takes stay readable. Old v5 or unsafe history is not repaired or promoted into reusable Sampling v6 history; start a new Run when required. | +| Unknown MODEL wrappers | Fresh generation remains available; reuse requires observable, matching contracts. No blanket approval by class name or version. | +| Attention / Spectrum | Finite audited profiles support Core PyTorch/Kitchen and specified Spectrum implementations, including upstream v0.2.29. [Exact scope](docs/ISSUE31_SAMPLING_SIGNATURES.md). | +| KJ Model Preview Override | Only the pinned KJ/Core source profile: **VAE / Audio VAE disconnected, Tiny VAE `none`**, Core NoPreviews/Auto/Latent2RGB. KJNodes 1.5.2 alone is insufficient; the audited local source differs from upstream. [Supported hashes](docs/KJ_PREVIEW_OVERRIDE_V398.md). | + +Published **v3.9.9** passed new-Run, separate-Queue Chunk 1→2 GPU continuation +with **PyTorch + KJ** and **Comfy Kitchen INT8 + audited local Spectrum + KJ**. +Preview images remained held, original Chunk 1 data/contracts stayed unchanged, +and all four video/audio outputs decoded successfully. +These were 512×512, 24fps, 2×5-second, 8-step Turbo Core-API tests on an RTX 5060 Ti 16 GB. +They do not establish support for other KJ/Spectrum builds or preview-VAE paths. +[Full conditions and verification limits](docs/RELEASE_NOTES_V398.md). + +## Known limitations + +- **Reference Plan Inspector:** stale display after repeated Sampler configuration + remains a known issue; the two expected test failures do not mean it is fixed. +- **Mobile:** reduced UI work is verified; mobile freeze resolution is unconfirmed. +- **Resources:** nine high-resolution References or long final decoding can exhaust + VRAM/RAM. Disk-backed Finalize does not prevent allocations in other nodes. +- **Refinement:** high-ratio latent resize can cause artifacts; longer Hi-Res/Hybrid + cases and doubled outlines remain limited or under investigation ([Issue #27](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/27)). +- **Continuity:** exact identity, motion, lip sync and frame-by-frame reproduction + are not guaranteed. Some long continuations have known quality limits. +- **Acceptance:** browser Run-button/save-reload, the reporter's exact full graph, + other provider builds, subjective quality and speed/VRAM benchmarks were not + established by the targeted v3.9.9 GPU recheck. + +## Detailed documentation + +| Topic | Read more | +| --- | --- | +| Setup, screenshots and complete controls | [Illustrated user guide](docs/USER_GUIDE.md) | +| Current Reference Images | [V3.9 Reference guide](docs/V39_REFERENCE_INLINE.md) | +| Updating from V3.8X2 | [Migration and retained contracts](docs/V38_RELEASE_AND_MIGRATION.md) | +| History, earlier screenshots and implementation notes | [README history](docs/README_HISTORY.md) / [CHANGELOG](CHANGELOG.md) | +| Current release and GPU verification | [v3.9.9 notes](docs/RELEASE_NOTES_V398.md) / [package validation](PACKAGE_VALIDATION.txt) | +| Bug reports | [GitHub Issues](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues) — include the workflow, status text and `auto-resume disabled:` reason | + +Historical Releases and tags remain available. Documentation updates on `main` +do not replace the published v3.9.9 tag or package assets. +This condensed README changes documentation only; public nodes, official +workflow JSON/ZIP files, Sampling v6 and existing Runs/Takes are unchanged. + +## License + +[MIT](LICENSE). Third-party components retain their own [licenses and notices](THIRD_PARTY_NOTICES.md). diff --git a/README_JA.md b/README_JA.md new file mode 100644 index 0000000..888cf11 --- /dev/null +++ b/README_JA.md @@ -0,0 +1,279 @@ +# ComfyUI-H3-Continuum 3.9.10 — V3.9 + +MiniMax H3の映像・音声をチャンクに分けて生成し、結果を確認しながら +一部分の再生成や、互換性のある保存Takeからの継続ができます。 +**現在のWorkflowはV3.9です。V3.8X2も別のWorkflowとして保持しています。** + +[v3.9.10本体ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.10/ComfyUI-H3-Continuum-v3.9.10.zip) · +[Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.10) · +[チェックサム](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.10/SHA256SUMS.txt) · +[English](README.md) + +## 3.9.10 — Spectrum再利用とReference割り当て表示 + +SpectrumソースのCRLF改行でSampling v6 read-onlyになる判定を修正し、対応済みの +2種類の実装にPython 3.12用の監査済みプロファイルを追加しました。既存3.13の +プロファイルと正常なTakeの識別値は維持します。未知のコード・追加状態は拒否します。 + +外部Chunks接続時のReference Imagesは、古いWidget値の代わりに割り当て範囲1–16を +表示します。生成チャンク数や保存済みの選択内容は変更しません。 + +CPU/JS:**2,096成功・2スキップ・既知の期待失敗2件**。Python 3.12の対象試験、 +保存・継続と実JS照合も成功。Reference表示は実ブラウザー確認済みです。 +**3.9.10としての新たなGPU受入は未実施です。** + +更新後はComfyUIを再起動し、ブラウザーを更新してください。既存WorkflowとTakeは +そのまま保持できます。新しく対応した構成は新規Runで確認してください。 +旧unsafe/v5 Takeは閲覧用として保持し、自動変換しません。[詳細と制約](docs/RELEASE_NOTES_V3910.md)。 + +## 3.9.9の修正と検証 + +確認済みの公開KJ Preview Overrideで、新規Runまでv6 read-onlyとなる対応漏れを修正。 +再利用不可のTakeには保存時の理由を表示します。安全判定・既存Take・Workflowは維持します。 + +CPU **2,075成功/2スキップ/既知の期待失敗2件**。公開KJのGPU継続と、既設KJを維持した +実ブラウザーのReviewボタン・Continue From Here→Runを確認し、**Chunk 1保持・Chunk 2生成 +(1再利用・1新規生成)**に成功。4本の映像・音声を全デコードしました。 + +更新後は再起動とブラウザー強制再読込を行ってください。Preview Override側の任意VAE/Audio VAEは +未接続、Tiny VAEはnoneにし、通常の生成用VAEは維持してください。旧unsafe/v5 Takeは保持されますが +更新だけで再利用可能にはなりません。設定修正後は新Runで開始します。配線変更・Take削除は不要です。 +[検証条件と制約・Release notes](docs/RELEASE_NOTES_V399.md)。 + +## 基本操作を動画で見る + +[H3 Continuum V3.9 — Chunks, Prompts & Reference Images](https://youtu.be/AKJxBXaiG6Q) — **12分26秒**、英語ナレーション・字幕・チャプター付き。 + +First Image、チャンク設定、Review・Takeの再試行、Timeline Prompt、Reference Imagesを説明しています。 +画面写真と詳しい操作は[日本語の詳細ガイド](docs/USER_GUIDE_JA.md)を参照してください。 + +## インストール・更新 + +Continuumは`ComfyUI/custom_nodes`内に1つだけ配置してください。 +モデル・外部カスタムノードは本体ZIPに含まれません。 + +### 固定バージョンのZIPを使う + +1. 冒頭の**本体ZIP**を取得し、必要に応じてチェックサムを確認します。 +2. パッケージフォルダーを`ComfyUI/custom_nodes`へ展開します。 + 更新時は既存パッケージを置き換え、重複したコピーを残さないでください。 +3. ComfyUIを再起動し、ブラウザーを**Ctrl+F5**で強制再読込します。 +4. 下記のV3.9 Workflowを開き、導入したモデルと入力ファイルを選びます。 + +保存済みWorkflow、Run/Takeフォルダー、設定は保持してください。 +**Workflow ZIPはグラフの配布物であり、カスタムノードのインストーラーではありません。** + +### Git・ComfyUI-Managerを使う + +Gitで新規導入する場合: + +```bash +cd ComfyUI/custom_nodes +git clone https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum.git +``` + +既存のGit導入を更新する場合: + +```bash +cd ComfyUI/custom_nodes/ComfyUI-H3-Continuum +git pull --ff-only origin main +``` + +Managerで導入した場合は、そのコピーをManagerから更新してください。 +更新後はComfyUIを再起動し、ブラウザーを再読込します。 +Registry審査・Managerが選択する版・GitHub Releaseの状態は別です。 +正確なv3.9.9を使う場合は、冒頭の固定版リンクを参照してください。 + +パッケージの宣言上の最低要件は**Python 3.10以降・ComfyUI 0.32.0以降**です。 +この最低要件の宣言だけで、すべてのCore版でのGPU動作確認を意味するものではありません。 + +## 必要なモデル・拡張機能 + +重みは[Comfy-Org / MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3)から別途取得します。 +次の表は**保存されているV3.9テンプレートの選択値**です。配置先は`ComfyUI/`からの相対パスです。 + +| 用途 | ファイル・ダウンロード | 配置先 | +| --- | --- | --- | +| H3生成モデル | [minimax_h3_fl2va_pruned_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors) | `models/diffusion_models/` | +| テキストエンコーダー | [qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors) | `models/text_encoders/` | +| 動画VAE | [minimax_h3_video_vae_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_int8_convrot.safetensors) | `models/vae/` | +| 音声VAE | [minimax_h3_audio_vae_fp32.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors) | `models/vae/` | + +各Loaderで取得したファイルを選び、テキストエンコーダーのtypeは`minimax`にします。 +保存済みの`minimax/`サブフォルダーが環境と異なる場合は、実際の配置先から選び直してください。 +動画VAEの選択値はINT8版ですが、旧ダウンロード情報にFP16版が残っています。 +実際に導入したVAEを選んでください。[別のモデル形式](https://huggingface.co/Comfy-Org/MiniMax-H3/tree/main/diffusion_models)は生成用途とLoaderの対応に合わせて選択します。 +量子化形式には対応するCore・GPUが必要であり、上表は全ハードウェアでの動作保証ではありません。 + +保持しているV3.8X2グラフでは、DasiwaMinimaxH3_dasiwaHybridV2_int8.safetensorsと +[FP16動画VAE](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors)が選択されています。 +このカスタムモデルは同梱されず、上表の配布元から取得するものでもありません。導入済みの互換H3モデルを選び直してください。 + +配布グラフ全体を使う場合は、次の外部ノードも必要です。 + +- [Spectrum](https://github.com/xmarre/ComfyUI-Spectrum-MiniMax-H3):任意の高速化機能。保存時は無効です。全バージョンでのTake再利用を保証しません。 +- [KJNodes](https://github.com/kijai/ComfyUI-KJNodes):テンプレートのH3 SageAttentionパッチを提供します。必要なAttention依存関係を導入するか、利用できないパッチを置換・Bypassしてください。 +- [rgthree-comfy](https://github.com/rgthree/rgthree-comfy):`Power Lora Loader (rgthree)`を提供します。 + +**Turboを使う場合のみ:** 用途に合う[LightX2V Turbo LoRA](https://huggingface.co/lightx2v/Minimax-h3-Turbo)を1つ選び、`models/loras/`へ配置します。 +標準WorkflowではLoRAはすべて保存時に無効です。別のTurbo Sampleでは1つ有効です。配布元のSampler・Steps指定に従ってください。 +[ファイル一覧と設定の詳細](docs/USER_GUIDE.md#turbo-lora-and-spectrum-in-supplied-workflows)も参照できます。Continuumがこれらを自動導入することはありません。 + +## Workflowを選ぶ + +| Workflow | ダウンロード | 用途 | +| --- | --- | --- | +| **V3.9** | [JSON](examples/workflows/MiniMax_H3_Continuum_V39.json) / [ZIP](examples/workflows/MiniMax_H3_Continuum_V39.zip) | 現行Samplerと固定Reference Imageスロット | +| **V3.9 Turbo Sample** | [JSON](examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.json) / [ZIP](examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.zip) | Auto・2×8秒のダンスとタイ語ラップ。[設定・制約](docs/TURBO_SAMPLE.md) | +| **V3.8X2** | [JSON](examples/workflows/MiniMax_H3_Continuum_V38X2.json) / [統合ZIP](examples/workflows/MiniMax_H3_Continuum_V38X2.zip) | 既存Workflow用。統合ZIPにはV3.9も収録 | + +ComfyUIの**Templates → ComfyUI-H3-Continuum**からも開けます。 +V3.8X2のReference配線や保存履歴はV3.9へ自動変換されません。 +元のグラフを保持し、新しい作業は専用のV3.9グラフで開始してください。 +不足する外部ノードは導入・置換・削除してからQueueします。ComfyUI-Easy-Useは不要です。 + +V3.9テンプレートでは、自分の**First Image**を選んでください。 +Reference Loader 1〜6はBypass、チャンク別割り当てはすべてOFFで保存されています。 +Image 7〜9を使う場合はLoaderを追加・接続します。 +[テンプレートの詳しい説明](docs/USER_GUIDE_JA.md#最初にv39とv38x2)。 + +## 最初の生成 + +長い動画を作る前に、短い2チャンクで確認してください。 + +1. H3モデル、テキストエンコーダー、Sampler、対応VAEを選びます。 +2. `Sequence Prompt`を記入します。全チャンク共通なら`Prompt Format = Fixed`を使います。 +3. 画像から作る場合は`Size Source = First Image`と解像度を選びます。 + 初期値は**Draft — 0.30 MP**です。幅・高さを直接指定する場合は`Manual`にします。 +4. `Chunks`を**2**、`Seconds per Chunk`を**5**にし、`Total Length`が**10秒**になることを確認します。 +5. `Base Seed`を維持し、`Control After Generate`を`fixed`にします。 +6. `Run`を`Review Each Chunk`、`Progress`を`On — Resume and Takes available`にします。 +7. `Regenerate From = Auto`のまま`Ready to Queue`を読み、ComfyUI上端の**Run/実行ボタン**を押します。 + +`Chunks`は最終目標数であり、次のQueueで追加する数ではありません。 +Reviewでは通常1単位ずつ生成するため、この例の最初の出力は**5秒**です。 +Last ImageのTerminal Mergeでは、最後の2チャンクを1単位として扱う場合があります。 + +一括生成する場合は`Generate Full Video`を選びます。 +保存Take・互換Resume・後からの再生成を使う場合はProgressをONにしてください。 + +## 確認・継続・再試行 + +Review操作を選ぶだけでは生成は始まりません。選択後にComfyUIの +**Run/実行ボタン**(`Queue`)を押してください。 + +| 操作 | 次のQueueで行うこと | +| --- | --- | +| `Use it and continue` | 採用済みチャンクを保持して、次のReview単位を生成 | +| `Try this chunk again` | 前のチャンクを保持して、確認中の単位を別Takeで再生成 | +| `Use it and finish the rest` | 採用済みの部分を保持して、残りをまとめて生成 | +| `Back to Settings` / `Return to Review` | 設定と確認画面の移動。生成開始・Take削除は行わない | +| `Use This Take` | 保存結果を採用。新しいSamplingは行わないがDecode・保存は実行され得る | +| `Continue From Here` | 選んだTakeまでを保持した分岐を作り、次の単位を生成 | + +**2×5秒**の例では、Chunk 1を採用して継続すると、それを再利用してChunk 2を生成し、**10秒**の完成動画になります。 +完了後に現在の単位をRetryすると、別のChunk 2 Takeを作ります。 + +継続時はモデル・LoRA・Seed・Prompt・入力素材・寸法・Sampling設定・継続設定・Run識別を維持します。 +Seedを固定しただけで再利用が保証されるわけではありません。Statusと再利用数を確認してください。 +完成動画を延ばす場合は条件を維持し、`Regenerate From = Auto`のまま最終目標の`Chunks`を増やしてQueueします。 +[Review・Takeの詳細](docs/USER_GUIDE_JA.md#10秒を5秒ずつ確認しながら作る)。 + +## 指定チャンクから再生成する + +1. 設定を開き、**Regenerate From = Chunk N**を選びます。 +2. ProgressをONにし、前のチャンクの生成条件を維持します。 +3. N以降をまとめて作るなら`Generate Full Video`、次の単位で確認するなら`Review Each Chunk`を選びます。 +4. **Variation Nonce**を指定し、ComfyUIの実行ボタンを押します。 +5. 通常の継続・Take操作へ戻る前に、`Regenerate From = Auto`へ戻します。 + +Nonce **0**は、再生成完了後の再実行で別Variationを自動選択します。 +互換性のある中断処理では、同じVariationをResumeします。 +正のNonceはVariationを固定します。同じ条件・同じ正数を繰り返しても別結果の要求にはなりません。 +ReviewのRetryは、これとは別にVariationを管理します。 + +前の互換チャンク・以前のTakeは保持します。 +後続は変更後の継続に従うため、Full RunのChunk 2指定は、目標数によってChunk 2以降をまとめて再生成します。 +Terminal Mergeのpairは分割せず、チャンク内の任意フレームだけを変更することもできません。 +[再生成・互換性の詳細](docs/REGENERATE_INTENT_V396.md)。 + +## Prompt・Reference Images + +- **Fixed**は共通Prompt、**List**はチャンク別、**Timeline**は時間指定です。 +- V3.9のReference Imagesは**Image 1〜9の固定スロット**を持ち、Samplerへ1本で接続します。 + `All chunks`は共通使用、`Per chunk`は画像とチャンクの交点を選択します。 +- `@R1`は空きスロットに関係なくImage 1です。有効なReferenceだけをタグで参照してください。 + 実際のPicture番号は物理groupごとに解決されます。 +- First/Last Image、Reference Images、Reference Audio、Driving Audioは用途が異なります。 + Driving Audioは元の音声を保持し、Reference Audioは生成条件として使います。 +- Reference単独を試す場合はFirst/Last ImageをOFF、サイズをManualにします。 + 終端を結合するpairでは、両チャンクの有効なReferenceセットを揃えてください。 +- 大きな入力画像は事前に縮小してください。条件付けのresizeだけでは元画像の読込RAMは減りません。 + 割り当てを外した後も、映像の連続性により被写体が残る場合があります。 + +[Referenceガイド](docs/V39_REFERENCE_INLINE.md) · +[Prompt例(英語)](docs/README_HISTORY.md#prompt-formats) · +[V3.9用Promptスキル](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/tree/main/write-h3-v39-reference-prompts) · +[Prompt・スキル配布](docs/USER_GUIDE_JA.md#プロンプトスキルのダウンロード) + +## 保存・Decode・Second Pass + +ProgressはRun/Takeを保存します。任意の**Decode Cache Helper**はデコード結果のキャッシュであり、保存履歴とは別です。 +Finalizeで結合し、後段のCoreノードで動画を作成・保存します。 + +一時保存先は**設定 → H3Continuum → Storage → 一時ストレージを管理**で指定します。 +ComfyUIサーバー側のフォルダーで、FinalizeとDecode Cacheに適用します。 +Run/Takeや完成動画は移動しません。変更・削除は実行中/待機中Queueが空のときに行ってください。 +同じ物理ドライブ内のフォルダー移動だけでは、書き込みを別機器へ逃がせません。 +[保存方式・削除・制約](docs/TEMPORARY_STORAGE_JA.md)。 + +**Second Pass**は外部のRefinement・拡大処理へ渡す任意の機能です。 +チャンク継続や通常のReview Retryとは別で、V3.8X2標準WorkflowにHi-Res Fixは不要です。 +V3.8ではReview途中のpartial sequenceに対応しません。V3.9は検証済みの実出力groupの条件を継承できます。 +[外部連携の仕様](docs/V38_OPEN_INTEGRATION_CONTRACT.md)。 + +## 互換性・v3.9.9の対応範囲 + +| 対象 | 対応条件 | +| --- | --- | +| 保存履歴 | 既存Takeは読取可能です。旧v5・unsafe履歴をSampling v6の再利用可能履歴へ自動修復・昇格しません。必要に応じて新Runを使います。 | +| 未知のMODEL wrapper | 新規生成は可能です。再利用には生成契約の完全な観測と一致が必要で、名前・版だけで許可しません。 | +| Attention/Spectrum | Core PyTorch/Kitchenと、公式v0.2.29を含む指定Spectrum実装の監査済みプロファイルに対応します。[厳密な範囲](docs/ISSUE31_SAMPLING_SIGNATURES.md)。 | +| KJ Model Preview Override | 固定したKJ/Core実装のみ。**VAE/Audio VAE未接続・Tiny VAE `none`**、Core NoPreviews/Auto/Latent2RGB。監査済みローカルソースはupstreamと異なり、KJNodes 1.5.2全般の合格ではありません。[対応ハッシュ](docs/KJ_PREVIEW_OVERRIDE_V398.md)。 | + +公開版**v3.9.9**は、新Run・別QueueでのChunk 1→2継続を +**PyTorch+KJ**と**Comfy Kitchen INT8+監査済みローカルSpectrum+KJ**で確認しています。 +プレビュー画像を保持し、元Chunk 1と契約を維持、映像・音声4本は全デコード成功でした。 +条件はRTX 5060 Ti 16 GB、512×512、24fps、2×5秒、8-step Turbo、Core API経由です。 +別のKJ/Spectrum実装やPreview VAE接続を合格とするものではありません。 +[詳しい条件・未確認範囲](docs/RELEASE_NOTES_V398.md)。 + +## 既知の問題・制約 + +- **Reference Plan Inspector:** Samplerの再configure後に古い表示が残る問題は未修正です。 + 2件の期待失敗は修正済みを意味しません。 +- **モバイル:** UI処理の削減は確認していますが、フリーズ解消は未確認です。 +- **メモリ:** 大きなReference画像9枚や長尺の最終DecodeでVRAM/RAM不足になる場合があります。 + Disk-backed Finalizeでも、他ノードの大きなメモリ確保は防げません。 +- **Refinement:** 高倍率のLatent Resizeでアーティファクトが残る場合があります。 + 長いHi-Res/Hybrid条件と二重輪郭には制約・調査中の事項があります([Issue #27](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/27))。 +- **連続性:** 完全な同一性・動作・口の同期・元動画の全フレーム再現は保証しません。長尺には既知の品質制約もあります。 +- **検証範囲:** ブラウザー実行/保存再読込、報告者の完全一致グラフ、他実装、主観画質・音質、速度・VRAM比較は、今回の限定GPU再検証で確立していません。 + +## 詳細資料 + +| 内容 | 資料 | +| --- | --- | +| 導入・画像・詳細操作 | [日本語の詳細ガイド](docs/USER_GUIDE_JA.md) / [英語の図解ガイド](docs/USER_GUIDE.md) | +| 現行Reference Images | [V3.9 Referenceガイド](docs/V39_REFERENCE_INLINE.md) | +| V3.8X2からの更新 | [移行・保持する仕様](docs/V38_RELEASE_AND_MIGRATION.md) | +| 更新履歴・旧版画像・実装説明 | [日本語の履歴資料](docs/README_HISTORY_JA.md) / [CHANGELOG](CHANGELOG.md) | +| 最新版の検証記録 | [v3.9.9詳細](docs/RELEASE_NOTES_V398.md) / [配布検証](PACKAGE_VALIDATION.txt) | +| 不具合報告 | [GitHub Issues](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues) — Workflow、Status、`auto-resume disabled:`の理由を添えてください | + +過去のRelease・タグは保持しています。mainの文書更新で、公開済みv3.9.9のタグや配布物を置き換えることはありません。 +この整理は文書のみであり、公開ノード・公式Workflow JSON/ZIP・Sampling v6・既存Run/Takeは変更しません。 + +## ライセンス + +[MIT](LICENSE)。外部コンポーネントの[ライセンス・表記](THIRD_PARTY_NOTICES.md)も参照してください。 diff --git a/REGISTRY_MANIFEST.sha256 b/REGISTRY_MANIFEST.sha256 new file mode 100644 index 0000000..c2c3a1b --- /dev/null +++ b/REGISTRY_MANIFEST.sha256 @@ -0,0 +1,171 @@ +ce319f06573b0b53709bae56d9b9ee0b3f4180eba961d7cbbac9e01e403a589e .comfyignore +fa9c99303d1a960db12ae2191cffcfbffe82e25e144eb6d16775c840fec4d573 CHANGELOG.md +fe29dafce4a6c5229e7e3aef1f946afba2346015702ed8cdda9927a80b6e65f3 INSTALL_JA.txt +c44abbd789091bbead01250dbd5453e60539d2665c69f162335fdbd2936019c2 LICENSE +656e417909695c5dcabf134b6f3240c44bd02bb2e33881b5247efd345cb932c2 PACKAGE_INFO_JA.txt +57fa9ad77ab5b1d2da6ec43f7ba19118e253cb4a959b758b246755471df15591 PACKAGE_VALIDATION.txt +da45eeb9f4950777bb63974cbf1d77856b287b05e3466be63571f2ad4561f12e README.md +eed26374c7de5e9e19396836e0270fb46f37f3302dfd6f5533c8d703741cbf0f README_JA.md +0dbe8a5f71c92f3523942343b40787104c8b96781461ba3ba82934a6bdbe5834 THIRD_PARTY_NOTICES.md +02ea0951ce2e7ac2d5f3bbf02c596a9f9ea8adcdf2a113dca4c8db3cc19c36b3 __init__.py +7fd122ff08232e114a6ef32171217ea6324e980eb7257d560c466b8177a3d3ec audio_compat.py +3e7a1ec3380be2a222c8a3e9b8a6bb184d3a760478957f48186438f15235bdba branch_provenance.py +c1ed3a0342730450958fb62fd21854a876991f3a4aefb24e5383b0c73cd2312d compatibility.py +4d62508f900af9f93ec320afb900c27cc9a300d1e0c0a94714258498c5ede5fc conditioning.py +294251c40b1be2e296c5e5f627b68091d087bdb39179e3a0c3cebd2ee8a344f5 constants.py +cf255b8a635471229aebadc926acfdcacb65963adbcb9e0f9505d04eada2c6b5 continuation.py +51f1d211dfc36602a006fe5d14f4ebb08182694ad464aa9176340ff7e2bd1e41 decode_cache/__init__.py +647f55a9253976c059502dec13c7f64d7f7df1ecbe619d543c973f92f7f36776 decode_cache/identity.py +c2b44e3caed54c894556af547cd800328ed287a8cc90e0b5031f34bd6bd2224b decode_cache/nodes.py +986cc7a3dc153ff1b41c427a86756e505ea6e17d8b065df1ab15b8ff904f090e decode_cache/service.py +630ba3d2183964444c4341a8df507f7af61dac6a8e5aaca8cf387a5376d49f27 decode_cache/store.py +44377cbac9131daa3d1c88ccc154c325257767f4c52f3181f61525c7eae00531 docs/ISSUE31_SAMPLING_SIGNATURES.md +b7ecd069d803cae7f86ad0ab89df144b5a5b157c168c73ee2962b47234d4ec0c docs/KJ_PREVIEW_OVERRIDE_V398.md +77c225d1e49e06d00b04acf811177c9a4317e792d4bfe92df539de99420c3563 docs/LOADER_PERSISTENCE_REPAIR.md +b30fe4640359bebfcdeeb94dbeec94cb346c8e3646509e1dcdb87d7021ba04ff docs/N2B_BRANCH_PROVENANCE_CONTRACT_V1.md +bce0337cce903938b6b8df7f52c5c67e2d0ee14599c693687e4180debc606db2 docs/README_HISTORY.md +9ee92cdd3783e61e3bc71b99d5c0df7dd0a9f129c8a6b37701d7001a4f745e77 docs/README_HISTORY_JA.md +cebf8462681796df807987d8afef2c78abb0953b54b2f5dcd12d583bde0ec806 docs/REGENERATE_INTENT_V396.md +bbe74f802339b758d8874900e2617dd1a46844a58e0f872d95f34b9b40ba20ed docs/RELEASE_NOTES_V3910.md +49d02b7650d4a542307c9d4189dc217ba1db54c4eac959cc2e2a1ce57b501ae3 docs/RELEASE_NOTES_V399.md +91ee4adf828415e84e2f413e3a4e20dc16356f139e8ef1f6756f584cad04f3c8 docs/REVIEW_RESTORATION_REPAIR.md +4546ffb0cbaf270c52d1900908e46bfa2d3c0e2d0035ae5694680890e89aedd8 docs/TEMPORARY_STORAGE_JA.md +4c33d223a77c78e5c74946aa43f537031fd5aa415388d46fed975f9b07d53c3e docs/TIMELINE_VIDEO_EXPERIMENTAL.md +00f88cb57be5e79d9c881671bd13fa154d91843f4c1ae2f2451ee2a73841271d docs/TURBO_SAMPLE.md +92bee3f3c0c340356abdd06b4993fbf2e98cbe57e931f85cbfd11da0f7f87809 docs/USER_GUIDE.md +c3f0753bdae1c54044a6e9ad5770c45e85a306ac9c15cd363790f8625c865f94 docs/USER_GUIDE_JA.md +717577efdec5e61600504b77bc2a099ef022c4fe2333e936c3310c7182d8f2a0 docs/V38_OPEN_INTEGRATION_CONTRACT.md +3496636f645b39336fd95a4f21a562810c56f8e1d92fa6eeb1e4b0786ad11b37 docs/V38_RELEASE_AND_MIGRATION.md +4d17d74107e78404b190fec9494b443249e5aff4f58670c50d54dfdfb064334b docs/V39_REFERENCE_INLINE.md +a4705ae4b5730d2b43e29fe125a64196a031f14bcfa8702663a17aa89f683d89 docs/images/v39/v38x2-feature-summary-historical.png +ae3338f720fc22c05ab4a661c68f9c5dcb20d751a835f0e60d4738333be48e63 docs/images/v39/v39-feature-summary.png +84304d318eea6db3921d366647c6fff9ad488e8c4cfef6ca00988bfd2a6504bc docs/images/v39/v39-reference-images-node.png +b765b37502b012b462606c4d2dcef43f13950e5fd682824fb18cdb19d542c121 docs/images/v39/v39-sampler-node.png +a8c56379abe23b98c8ceed6ea5e60a02ecc3c6d9636232d7371615f4eb306e81 driving_audio.py +1b2212b7511fac0b63dfa6e9f006cd8644c72d08c6945ab0e4ed0e722c97a730 examples/MiniMax_H3_Continuum_V38X2.json +40199d2500174ba51c040d6983f88d74c4bfbaf19191e728818e6b969b869b0f examples/MiniMax_H3_Continuum_V39.json +705d1141842863b82abe26a8bb015d45eff3bc81ca5bdc53e58b464b743173c0 examples/MiniMax_H3_Continuum_V39_Turbo_Sample.json +1b2212b7511fac0b63dfa6e9f006cd8644c72d08c6945ab0e4ed0e722c97a730 examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.json +4fa8680037595f55f97715f9a2f63d28f11328a87f5c33c316cc4b0ae817e668 examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.zip +1b2212b7511fac0b63dfa6e9f006cd8644c72d08c6945ab0e4ed0e722c97a730 examples/workflows/MiniMax_H3_Continuum_V38X2.json +f3d8dd0f0872afdd97fe24e58b5cb74190d03e3583a66a5d013fee41f6894cda examples/workflows/MiniMax_H3_Continuum_V38X2.zip +40199d2500174ba51c040d6983f88d74c4bfbaf19191e728818e6b969b869b0f examples/workflows/MiniMax_H3_Continuum_V39.json +f69b734b5bf19fb6db78bbe4737db0bef69c2a6cef2710f28767882ef35ec847 examples/workflows/MiniMax_H3_Continuum_V39.zip +67e81fbdd4c21a8d64fb3bedb5e495698fc8df5cdd9540c8ef5476caa76119c0 examples/workflows/MiniMax_H3_Continuum_V39_Reference_Inline.json +65689fd800a60ab6939bea020d630616b03a453cf6d48022cd1361992a765abc examples/workflows/MiniMax_H3_Continuum_V39_Reference_Inline.zip +705d1141842863b82abe26a8bb015d45eff3bc81ca5bdc53e58b464b743173c0 examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.json +2096e583c6805c0f6f54f76a81e0ed1246325a16d189afba0423fa05e2c19599 examples/workflows/MiniMax_H3_Continuum_V39_Turbo_Sample.zip +a344da581b3c2d77558a460d2b5256e6c9a63b5008c98b52999ceda2e0b2b767 graph_contract.py +bf0556b076dbaca67c3405a7d86c333ca7fdf3d4381529b0a9edc906519943e9 guide_timeline.py +1473c384082700f422f2cf9141e0d012d87d2b43ac8a8a433edc7172e6ca7bf6 hardening.py +23bddc5a41d2e997f4fbf789c7128e32f4f197f1405b2aef1c76794f57482446 install_windows.bat +12bf572a65d57bbc9a72b459f29691c2057768f4fb7a74f9db87f4775e995bda install_windows.ps1 +139546ec1ff459945aa89b50f74b5ee69f62a0e07bbddfcae06936091a409a83 layout_adapter.py +d3e213005670e94dee33a240349a4d245cf9a3b6602099f11ed30cf5a32e2c21 media.py +7ad8ea8859b666cb2771f19795ca8c04ab7aafb5c20cff4e80ad17295429ded3 metadata.ini +d7e613e2cc24110213b477bdd1d390234a41f2af3b1ed006a809066d71d658eb model_patch.py +ecb65a5dbf9e1683c5a3cadd82daaea3a90a2b776f6ebabfbe8b07aa93177422 nodes.py +6ba95639b52219fc37b4f51d52cf6879f75c7418a8d8815b3725b3f201a4f1e6 preview_override_provider_public_v1.json +d4cf474cbd1a98867075c45b53b2ab4d47752bb11611c916bac2a6831ca82c0b preview_override_provider_v1.json +3b7e18bd2ba50fd6c8fe71c58ca79a0bf752f362ed78b7357ce1998efc934b4b preview_override_signature.py +b470819f16fc179654dc1600721a63f0a5a14e777f4c004d574b499da6886ba5 public_api.py +dc56e02aa94561076bee8a836c268c0e45a8b5248f3b40108bcb740252965ec6 pyproject.toml +caa572ea65dafc86850db99f52433d360ac9c4ab36b854292a3dccd5ba32dc76 reference.py +9539f5cb9fb695841073fa58d9246d5a7410bc372f6cd340378185004dfaf978 reference_audio.py +5cb9bfd6810d3a93a637451980a4cc5c1436e5b84cfd1c7a111cd192e876351b reference_video.py +467edb2c188d024c2392aafe7423b3f835e16b10f570da76e2f7dda8a962c1c5 requirements.txt +985e7647bd2626af5df68ca702cf6ebb38cc09f3ffe0e239d3e8ef256352a5d8 run_storage.py +8f786989af6339be8cec8f2bf602da89344cf96202d29c980af5c07175eea3d3 spectrum_provider_v029.json +8cdf37b165eb68462dd7accf4a1fd63043f02479f9c377778fb9d4b6feaddcc6 spectrum_provider_v029_py312.json +0efc8d65cb6f59042457ad843fe1b77a4bee45e59d3732f5c83fbc75a31ae77b spectrum_provider_v1.json +71801234f6d2e08b660803194653821ca648e5b9a0d0c3f138bc7ccc6fe9d8c6 spectrum_provider_v1_py312.json +448213ae7670a6e231f833ca1901ca295eb5f6bfb8983ca37703bdb3fdf15f99 spectrum_signature.py +3a359a0467725cc292ccb4c86cf0c3d30bb6a114d01f5345792b09ea3d45441d state.py +6056de6c0a5322a8a51ba5a114538dca2ef67788cd1afba10fedd483b95f6afa state_io.py +660a9df28c7d62ad383cda5514c48c7efcb3576f435af023040116d533a56563 storage_api.py +517662b104ab1507bb927a39f25b8503f092c2ba4c39bc277379ba9d9e31b334 storage_fs.py +daeda14d9f5338835661e6738e074a3897d936cd847fe3e78519efabbcdfd0ef storage_memory.py +53da3ac42f4a66e8adc4560049c7c8e7ada82599be8c61f0573c234342ca36ed storage_settings.py +43ec4023dc06c941e5d7a04db4b8a89fb58d744ceb986f6eb9bf534d4fdcb87b temporal.py +f218e1a93c42467c6057c5c1d04d57cc82f6331b97ec9257e4e2dfb9c831876a timeline_video.py +0bc369ee428b8a5bc8e370a81fbab427262a1fdb7805a78c442c91698d65efba tools/__init__.py +2743597d289d9eb12334d1f5cdd630497e237814ddf48b17f6f68da299c01150 tools/p20_file_backed_tensor_poc.py +aab887698f4767722c833b9e55d605f9597ab0d823e32e8c5416f26a210cab1d tools/verify_runtime.py +ffd61f3f31c9b6430e1169007ee906c8f5000eba63ee44e2976ad4f5f1112671 v2/__init__.py +2003902c87bba09cfa143f5a9e4248f3a78c0fb544449a75fd40ba2cacb35655 v2/context_diagnostics.py +cde27db19b677e318cc4dfb4235be7a11429c19945f4083052a91fe117e74c82 v2/decoder.py +31012ec48a64c7467ad869c79982b7f8502c1785c33f51ee1b5727ce8b4243fb v2/diagnostics.py +2ebff4502381f675d06bba4b127b81bc3201e7579ddcd46fa5051cbdf3facd55 v2/h3_builder.py +fed4bc3cbff47bd0b16b6e99385680dc2253c745f30b7f686afafbb25f77efb1 v2/motion.py +de47a847ba27390df542086735a926d2812d77ad05b21ccc2e67633d4ffd11b0 v2/nodes.py +dc2b901789e4b024dbcbc4024c63eb8cbc564d77003ce8c35c0584851ba6dfbd v2/prompts.py +4080f1875b958b316e8e76852c4be5527a6de5c71d604e7642d3098478eb4a36 v2/sampling.py +2fd811794c4cc0f2ce050250b80484876454926b9f20fce6af3a97d728e5acdf v2/seam_guard.py +be0d8585fa57e0ccd6d02577bbc7299629e5bfe20e186747f571fccc98b67d7e v2/seam_types.py +12cfcc0c71ca087fe39a5ce2922959a62b2bb311bfe664f05c4480affce03708 v2/seeds.py +0de733c55bb61e69d48918eb5bc91318fc7cf3af383fb8937522ced9df018c47 v2/sequence.py +3ea23dbce6fcd217b6e81235855479b9b16fed7a21693cf156f9014555c4fe13 v2/session.py +13213e7709fd02bb810af3c18e4cd7d010f8cdfb121a4b99866a9f2333331089 v2/session_io.py +78c830cdf49034c6e3b2cd3da5914d5cfc5638eb63ec6dbd1bab13c50d6d4caf v3/__init__.py +ea2922567df90436dd6fa21696833a49dd0cbfeee142a5fedc3a7c303824d473 v3/adaptive_continuity.py +7815230e0a9540bc2edeaf818e940b7a1fdd817fd217ba8a4f9cb21aaf6a308e v3/assembly.py +70bec17d0dce1873b9bb4a51b09daf76833722ce89f005a22b029533a367e62b v3/assembly_v35.py +3f62fdb35e804f42d7cbf0828b4e87e69a2b99246c3f7467fe4fa698f1d803c9 v3/conditioning_bridge_nodes.py +6403c2bde3fc0b8cddef78567f663fddb48130628f4c9258fcd679c28fd0cf13 v3/driving_nodes.py +147a1d3fc5e09b323a4c3d13d30e432410389bae52b58330462375b9f307b791 v3/easy_audio_nodes.py +64b3e54b42e5156642f9957ccc374fd190d9fc7145637984e702534fb1bf80b4 v3/easy_image_nodes.py +ac859cb1a6e23b8ceae300d818c24fe1346b649a0d663017a9b3ac64aa96fca3 v3/easy_nodes.py +11e9daa45237392ac84ac1b54b3e8d5e980e77102779831824794a2a6675500b v3/easy_video_nodes.py +fe03ae0aa0dbbbe5f950bb7da34c18a078d841547992c374ea8beda5e3ad4715 v3/execution_planner.py +69bf49d0601edbe51730b9d409b1cde647feabde0bde561e416b0b472a27f4e6 v3/file_backed_buffer.py +cb53e1b104415032ec79b7e7aa1c99848e63da3eedaa251eeba102970e4172b5 v3/hires_fix_nodes.py +0858487e2b8c0060a856273ffc65d23165b2f253d183f506e27e419aea0da561 v3/latent_resize_nodes.py +5b8a54e67a608e757a505abc712effb33312636e3bb235835a485a506e80eeee v3/masked_continuation.py +91a868d2e392b018b01d0396209849c146b0d560ceb109fcdd87dea7f3d786aa v3/memory_action_policy.py +85477badd85281cfb77582fdd5b34228a9f6b68376f1b7743af2ca7d4b55aa36 v3/memory_attribution.py +4a48f74584d8ba3187e1accc31ae5acd3a2ffb9a491bb4f5ca43322db418e770 v3/node_facade.py +f2779a7453a84dec8b993cf044e69ec90da6fe86940a73181234b5000f5a8fa4 v3/nodes.py +a3411c7a295308d21623551cf4aeca5bef511eb5f063f31ac84032e3bdf4e16d v3/packed_row_planner.py +b7c4a35a5e73f528492bc6946b97b687c412f1d3f5f976c19d45c6dc20288c9a v3/pixel_vae_resize.py +7e8681bc1ca1633389e1af1ab14100338d23a047f96c59c19151acbbfae5c108 v3/plan.py +0e3dd69d6d2bec6d9491c2fb806c25f0c62629b1eb5f9e4f77ae8d27a5b19cfc v3/planning_types.py +ec71e3965f445756d54d005c0e27bea520e26dbac44e1ee3d3867bfe2da100b0 v3/ref_encode_cache.py +8b9c1cadf76e2c2425d6bf8a5783ced593a75f83a6b3f219a0f870b4c177e058 v3/reference_effective_plan.py +b64f4d3d9547ac00f8edb2155109476df5b8baea730fb94e8b5bfca6d11564e6 v3/reference_images_v39.py +4cc53a322960373949cc4e86a3102c04d3dce4dfac00f7cc62b01cf637775b62 v3/reference_plan_inspector.py +196d1b6f4a73c4fb8669e9c399ffae526bd74765f96fbf5340c780f4331e5594 v3/reference_routing.py +1366da221237853eddc9e4b724d4c28917868ea32607dd086361a84df052cfde v3/reference_runtime.py +4fb950bbdc7cdb55365fde373ee190af6cdf99f7ce4c54ef73b94f0aca4e1fda v3/reference_storage_contract.py +057921bf2b05e88c2cbc08c16e2ac483aea905b8a144fbadf4a21d6ae8252427 v3/reference_video.py +8d6cd0d7f507fd6c2d33667a6ea205e131d47571444abc94d529a98748a52517 v3/refine_context.py +519586b1b3f48c73d329f7a2aaa8a1bfc6fe31ac2ecb64f2b1fdf9c19026001a v3/refine_sampling.py +43f33f989eb70ca87b916fc33267732a2be3dd4581fc72af17c53f840b53ca49 v3/refine_schedule.py +afc149dc347b4c1502ddf32ea4229e633799cd34d2b2c6f98bf8a47910a9ef77 v3/refine_scope.py +ac373d9783d3a7bb968e8117843d246d6a090b66a819015348bff84e7747442c v3/refine_target.py +45ae91f13e8f71405d2f46045c8acc6eabbc9500c194cf5d0ba541bb085cd135 v3/refine_window.py +ae909b43ad17934c85881d3b0bea81e7b45c215eeb4d1f9f6a9ea72faf633ddf v3/reliability_v38.py +520a4ae849f83450855ce0023c582db5f91d29d8694a99aa8cf491351c837cff v3/resolution.py +3f03d8659305d2def41939205d21cdc6f35e32943274e87afe539d15b9635f9b v3/review_control.py +be6610fb9427546d9485a17015cfbdc8786715e6a0967f3e4c1dee35d8768385 v3/runtime_coordinator.py +d0a004aa3030c8adaa5ea788f9acc758fcb4a556f203f9a3f63fa4e4b5772063 v3/sampling_engine.py +97fcb293e0714c3a74b89f08e11c21dce45d25a3010948be1af9dc059d288187 v3/second_pass.py +452022107cea233aa84d5366ec4943ac5aadbd5fcf786113f1d542a641eac8f7 v3/second_pass_nodes.py +6e71e944313339f14e14e11cec682593159ca5542aa5d38095371da781d60224 v3/targeted_refine_sampling.py +5a044bdd70c546e4446e0c26af92aae88049ffa2e0e2d01b52033a83a3152d0e v3/targeted_second_pass.py +26b13614ba0a5fd8b192a106c9503b2f4b6fe7f9268a0fcb0f544979a49cab34 v3/video_adapter_nodes.py +2f07ace1462644f0bee553827b3571535de4d0c2d9ef847bd9e0ca73a4eed5cb v3/video_seam.py +4178dcfa442abdb06ce276e05bd7bcbaa6da28469ba5e57d3da3c663de4e4a9d v3/video_seam_v35.py +e124ded4bc8ddbf7b432c80d9900e40ee34214327bc6d83bd9fcf95cb55e7169 version.py +b935ad3bb729948bb3890ce967b0ecbef737a5e51ddfc47a0060fc20aa78dcda video_reference_modes.py +685df8e0fa23e3663ab31da1e4470cd19bdaa0dc16362cb606e874d9d6153c2c web/decode_cache_helper.js +fefafd4550511d80bd85ade46a0476f0cf3de13ddb94a908dff6b175f2a0e199 web/easy_bypass_toggle.js +092088096307d33ac016b82b4774ca38845a2ff2f63b97ac6da8c5aca4fea413 web/easy_load_audio.js +9db0aa94f634578a962d5db49e5c70356040a59df968e3a2e531b327a623081f web/easy_load_image.js +35090816ba7470a7255eb85d003b7e79f8a0065982dacf98a18f41c8ebff655d web/easy_load_video.js +0c68d13a1f8beb7431a32b471f8b0612c42fcc4bf72dd97611a025d158008662 web/easy_v38.js +933b5c675294a20ccec263916a1f4453d5b06a530d660ad8e56389d02d7f3023 web/project_id.js +ae59273f163a116ece418957c557c4fd8c3285e3c60814679b9361d2e6e6fa02 web/reference_audio_ui.js +46c9f75c9462bf26300332448f4431cca801f38964ca0a49bc2c50a88a7cc40f web/reference_image_ui.js +e99c74b78da1aa18be61cb053fa05e5abd106caa5eb5c814cf585786aae1e7f8 web/reference_images_v39.js +42afab2241d39c1b07ff33d8b1370d4b0f729d81e53b4c99f4bc54c6a96b8800 web/storage_settings.js +bf41bf08dd0504eb753fa5781e6c2b8c1285758ad4bb7309975bb9231d5f8e41 web/video_reference_mode_ui.js diff --git a/THIRD_PARTY_NOTICES.md b/THIRD_PARTY_NOTICES.md new file mode 100644 index 0000000..4426373 --- /dev/null +++ b/THIRD_PARTY_NOTICES.md @@ -0,0 +1,12 @@ +# Third-party notices + +H3 Continuum Join interoperates with, but does not bundle source code from: + +- ComfyUI native MiniMax H3 implementation +- ComfyUI-Spectrum-MiniMax-H3 +- ComfyUI-SolAttn_triton +- SageAttention / ComfyUI-KJNodes + +The package calls public ComfyUI sampling, VAE, ModelPatcher wrapper, NestedTensor, and native H3 payload/layout contracts. Optional accelerators must be installed separately and remain governed by their own licenses. + +This package was independently implemented. No source files from the optional accelerator repositories are included. diff --git a/WORKLOG.md b/WORKLOG.md new file mode 100644 index 0000000..350627f --- /dev/null +++ b/WORKLOG.md @@ -0,0 +1,189 @@ +# Public development log + +## 3.9.9 — KJ Preview Override / Review continuation + +Fix new-Run v6 read-only with the separately audited upstream KJ profile, and +show saved reasons when reuse is unavailable. Keep the original local profile, +Sampling v6 guards, old Takes and official workflow bytes. + +CPU: **2,075 passed / 2 skipped / 2 known expected failures**, plus overlapping +90-case JS coverage. Upstream KJ GPU and normal-ComfyUI browser Continue both +passed: **Chunk 1 preserved, Chunk 2 generated, one reused / one new**; all four +AV outputs fully decoded. Browser checks used the existing local KJ profile. +These are targeted repair checks, not general quality or all-provider acceptance. + +Restart ComfyUI and hard-refresh after updating. On Preview Override, disconnect +optional VAE/Audio VAE and set Tiny VAE `none`; retain generation VAEs. Correct +unsupported settings before starting a new Run. Old unsafe/v5 Takes remain +readable and are not automatically made reusable. No workflow rewiring or Take +deletion is required. [Full release notes](docs/RELEASE_NOTES_V399.md). + + +## 2026-10-16 — V3.9 Turbo Sample + +Add the user-supplied sample separately: Auto, 2 x 8s, multiline image-anchored +dance prompt, fixed background/camera, Thai rap and one enabled Turbo LoRA. +Preserve the supplied 6-step/Draft settings and document their unverified pairing +with the 4-step/768p LoRA. Fresh workflow/project IDs avoid user Run identity reuse; +remove misleading auto-download metadata for the saved custom model. Standard +workflows, existing Takes and runtime code stay unchanged. No new GPU validation, +runtime deployment, release or Registry publication is included. +Local CPU checks: 17 passed, zero CUDA initialization attempts. Workflow link +consistency, gallery/ZIP bytes, saved settings and both manifests were checked. + +## 2026-10-16 — Japanese README organization + +Align the Japanese README with the compact English entry point, including exact +model/VAE/text-encoder downloads, model folders, external node links and optional +Turbo guidance. Keep current compatibility and known limitations explicit. +Move all former Japanese detail and ten image references into USER_GUIDE_JA.md +and README_HISTORY_JA.md, preserving historical context and rebasing links. +Include these text guides in the Registry selection and update both manifests. +English documentation, product code, images, official workflows and existing +Runs/Takes remain unchanged. This is local documentation work, not a new release. + +Local checks: 5 existing tests passed, 95 relative file/heading links resolved, +all ten original image references and original detailed text were retained, and +both manifests matched (464 source entries / 162 Registry entries). The Japanese +README is 247 lines instead of 801. Its 18 external destinations match English. +No CUDA initialization, runtime deployment, commit or push was performed. + +## 2026-10-16 — condensed README and retained reference material + +Organize the English README as a short installation and operation entry point. +Keep the version heading, fixed-version downloads, getting-started video, Review / +regeneration steps, history safety and bounded v3.9.8 compatibility/known limits. +Move the former detailed controls and screenshots into docs/USER_GUIDE.md, and +release/legacy implementation material into docs/README_HISTORY.md; retain all +original sections and image assets, with rebased links and historical context. +Follow-up: list the template-selected H3 diffusion model, text encoder, Video/Audio +VAEs with provider downloads and ComfyUI model folders. Link Spectrum, KJNodes, +rgthree and optional task-matched LightX2V Turbo LoRAs. Explain the saved INT8 +Video VAE versus old FP16 download metadata without altering workflow bytes. + +Include the two text references in the Registry selection and update both source +manifests. Product code, Japanese README, official workflows, stored Takes, +release tags/assets and runtime deployment are unchanged by this documentation work. + +Local documentation checks: 5 existing tests passed; all 143 relative file/heading +links resolved, all 39 original image references were retained, and both manifests +matched (462 source entries / 160 Registry entries). README is 267 lines, down +from 1,574. No code, workflow, image or Japanese README changes were made. + +## 2026-10-16 — published v3.9.8 GPU recheck documentation + +Record a fresh GPU recheck after deploying and restarting the exact published tag `58fbd855`: PyTorch/KJ and Comfy Kitchen INT8 + audited local Spectrum/KJ each passed separate-Queue Chunk 1 to 2 continuation. KJ retained images, contracts and original Chunk 1 stayed equal, and all four AV outputs decoded. Existing Takes/workflows were hash-protected. README and Release notes distinguish this published-tag result from the earlier candidate gate and retain exact provider/no-preview-VAE/Tiny-none scope plus browser, quality and performance limits. This is a documentation-only main update; the original v3.9.8 tag and package assets remain unchanged. + +## 2026-10-16 — package 3.9.8 / KJ Preview Override Phase A + +- Audit the actual KJ Preview Override file, SHA-256 `d07c8e5ff7d7abd3e31acefe2cf29f618b5da5341b3b434e72ec2ac0953230b0`, rather than accepting KJNodes 1.5.2 by version alone. Pin sibling helpers and Core preview/dispatcher source, then compare loaded code/defaults/classes without importing or executing the provider during observation. +- Add a dedicated adapter only at `outer_sample/kj_preview_override`. Preserve all fixed wrapper settings and Core's key/list order. Ignore only the exact builtin transient `frames` list after the provider audit; preserve the generic no-KJ observer and all other reuse guards. +- CPU scope covers real-provider frame changes, reconstruction, fixed-setting/order changes, monkeypatch/source rejection, VAE/Tiny VAE exclusion, combined Attention/Spectrum identities, saved prefix bytes and old unsafe history. Full real-Core CPU gate: 1,995 passed / 3 skipped / 2 existing Inspector expected failures / zero unexpected failures and CUDA initialization attempts; the 48 dedicated KJ tests passed without skips. Normal ComfyUI GPU acceptance passed four Core-API Queues: PyTorch and Comfy Kitchen INT8 + the audited local Spectrum implementation, each with separate-Queue Chunk 1 to 2 continuation and held preview images. Both second Queues reused one chunk and generated one; original Chunk 1 bytes and contracts were unchanged. All four video/audio outputs fully decoded. See docs/KJ_PREVIEW_OVERRIDE_V398.md for exact support and limits. GitHub main/tag CI, Release asset integrity and Registry status are separate publication checks. + +## 2026-10-15 — package 3.9.7 / linked-input regeneration cache + +- Keep Full Run automatic variations reachable when Core leaves linked Chunks, Nonce, storage, regeneration boundary or generation mode unresolved during cache-key evaluation. +- Validate the actual chunk range at execution, and retain stable caching for known ordinary Full Run, storage OFF and directly specified fixed nonces. An unresolved external nonce conservatively reevaluates the node; Run Storage still owns prefix reuse. +- Keep Sampling Contract v6, Take protection, public schemas and official workflows unchanged. Add direct-input and real-Core cache regressions for both V3.8 and V3.9. + +## 2026-10-15 — package 3.9.6 / regeneration intent and automatic variations + +- Separate saved manual regeneration intent from the legacy stale Chunk 1 / nonce 0 state. Retain earlier unmarked later-chunk/fixed-nonce settings, and clear intent metadata when Auto or a one-shot restart consumes it. No widget order or API inputs change. +- Bypass Core output caching only for Full Run explicit regeneration with Run Storage on and automatic nonce 0. Run Storage remains the authority for compatible prefixes, interrupted variations and new completed variations; its Sampling v6 checks are unchanged. +- Add restored-workflow Queue cases for both public Samplers and a CPU-only real-Core cache lookup gate. Keep historical unsafe/v5 Takes read-only and official workflow bytes unchanged. Full CPU gate: 1,908 passed / 2 skipped / 2 known Inspector expected failures; zero unexpected failures and CUDA initialization attempts. Normal ComfyUI browser save/reload retained Chunk 2 / nonce 5 and Chunk 1 / nonce 0 after history loading. Five new-Run GPU queues passed continuation, Chunk 2 regeneration and two identical Full Run / nonce 0 submissions; earlier Chunk 1 bytes were preserved. GPU scope: PyTorch, Spectrum OFF, 512x512, 8-step Turbo; all five media outputs fully decoded. See docs/REGENERATE_INTENT_V396.md. GitHub main/tag CI, Release asset integrity and Registry status are verified separately during publication. + +## 2026-10-15 — package 3.9.5 / Issue #31 publication + +Main Linux/Windows CI passed. Correct the tag-presence probe to branch on the API exit status: a missing-ref error body must not be treated as an existing commit SHA. Product code and accepted GPU signatures are unchanged. + +Consolidate the Attention fix and finite Spectrum binding adapter on the exact 3.9.4 public source tree. Add a separate audited upstream Spectrum 0.2.29 profile without changing existing local-provider identities. Keep old unsafe history, public node IDs, contracts and all official workflows. CPU/GPU and publication checks are recorded separately in docs/ISSUE31_SAMPLING_SIGNATURES.md. + +## 2026-10-14 — Issue #31 Phase 1, unpublished + +- Final standard CPU validation: 1,777 passed / 1 skipped / 2 existing Inspector expected failures; no failures. All three standard commands recorded zero CUDA initialization attempts. Native PackedLayout, prompt planning and twelve-node registration checks passed. +- Limit production edits to the MODEL attention signature adapter in run_storage.py. Use already-loaded Core registries and implementation bindings; never import Core/Kitchen or execute an attention kernel during observation. +- Include both container functions and provider implementations/state without persistent addresses. Keep unrelated v6 signatures and every existing reuse guard; do not rewrite old unsafe manifests or raw files. +- Add real-Core CPU signature, clone/recreation, separate-process/hash-seed, mutation detection, prefix reuse and history immutability regressions. Kernel registration in CPU fixtures is structural evidence, not INT8 execution acceptance. +- GPU Review continuity, Spectrum/KJ execution, browser behavior and reporter-exact reproduction remain separate. No public node, official workflow, installed runtime, package version, commit or release changes. + +## 2026-10-14 — package 3.9.4 Registry packaging repair + +- Reconcile the actual Registry ZIP with its declared Manifest: include the Review restoration guide and exclude prompt-authoring source files. +- Add an exact archive file-set, CRC and hash gate before Registry upload using the same pinned publishing CLI. Preserve runtime code, public node IDs, workflows and saved data. +- Publish 3.9.4 as a separate version; keep prior GitHub tags and assets unchanged. Registry scan approval and real Manager installation remain independent of upload success. + + +## 2026-10-13 - 3.9.3 storage panel failure handling + +- Reproduce the initial panel failure with an empty HTTP 404. Startup logs identified a stale process retaining the configured address/port while the updated backend failed to bind; the new frontend could still be served from updated files. +- Keep Core's request address, base path and user/auth handling. Add explicit HTTP/invalid-response guidance, an initialized full-width storage candidate field, safe disabled mutation controls and manual recovery without reloading the page. Do not retry save or cleanup automatically. +- Keep the backend and generation contracts unchanged. Core 0.39.0 CPU regression passed 1,740 cases with 2 skips and 2 existing expected Inspector failures; CUDA attempts were zero. The frontend 1.53.10 browser gate passed empty-response diagnostics, manual recovery, save and cleanup at an alternate host/port. See PACKAGE_VALIDATION.txt for scope. + +## 2026-10-13 — package 3.9.3 temporary storage + +- Re-review the scratch storage proposal against actual Finalize, Decode Cache and Core settings lifecycles. Implement server-side settings with revision checks, queue coordination, immutable directory identities and no silent disk fallback. +- Add read-only status and explicit cleanup paths. Preserve live aliases and unknown owners; track detached helper disk entries using only metadata and weak leases. +- Add a ComfyUI settings panel without draw-loop scans or periodic polling. Keep existing workflow controls and node registration. +- Add CPU, real HTTP and filesystem regressions plus a Windows-focused CI storage gate. The real-Core CPU candidate passed 1,740 cases with 2 skips and the 2 existing Inspector expected failures, without CUDA initialization. Standard compile/runtime and actual Core/frontend settings save/reload/cleanup checks passed. All 25 JSON/ZIP assets retain their bytes; see PACKAGE_VALIDATION.txt for scope and platform limits. + +## 2026-10-13 — Issue #30 frontend hotfix, package 3.9.2 + +- Limit production changes to baseline-free Review snapshot skipping, one Reference selection notification per connected Sampler, and dimension checks before resizing the Reference helper. Align package metadata and current public guidance with 3.9.2; retain 3.9.1 records as history. +- Keep Review drawing-time upstream monitoring when a baseline exists and fresh synchronous Take/Queue checks. Preserve native callbacks, immediate saved assignment values, public widget/socket order and official workflow bytes. +- Add frontend regression coverage for upstream widget/link edits, rapid Reference selections, All/Off with future assignments, helper reconfiguration and independent helpers/Samplers. Track the independent Inspector reconfigure bug at 2 and 100 passes with explicit expected failures; one-pass refresh must succeed. +- Browser checks use isolated Chromium DOM and graph fixtures without generation. Full ComfyUI canvas, mobile performance and live workflow acceptance remain separate checks. +- Validation: 1,703 CPU/JavaScript regression cases passed, 2 skipped and 2 expected failures for the independent Inspector bug; zero CUDA initialization attempts. Offline runtime planning and twelve-node registration passed. In isolated Chromium, 600 baseline-free draw callbacks made zero snapshots (previously 600); 30 connected Reference checks notified the Sampler 30 times (previously 60), with zero same-dimension resize calls (previously 30). DOM rebuilds remain, and mobile freeze resolution is unconfirmed. + +- Final 3.9.2 candidate CPU gate passed 1,703 cases with 2 skips and 2 explicit Inspector expected failures. The actual Core/frontend/LiteGraph browser smoke passed checkbox interactions, Prompt/link/Bypass edit/restore and native assignment graph reload without submitting a prompt. This desktop headless/model-free check is separate from UI file-save, mobile hardware and stored-Take acceptance. + +## 2026-10-04 — V3.9 stability patch, package 3.9.1 + +- Preserve Second Pass audio/mixed-keyframe conditions and inherit the conditioning and References belonging to the selected Review output. +- Strengthen Take compatibility with Sampling Contract v6 / Graph Contract v4. Retain old v5 history without adopting its prefixes into new generation. +- Align Review Driving Audio with the original physical-group time range. Prefer Core audio resampling and document the same-object VAE mutation/cache limitation. +- Keep public V3.9 node IDs, official workflows and stored State/Session formats unchanged. LoRA Plan is not part of this patch. +- Full CPU validation: 1,696 passed / 1 skipped / no failures, with zero CUDA initialization attempts. Offline layout/planning and twelve-node registration passed. +- Targeted Review Second Pass and integrated generation/reuse GPU checks passed. Browser acceptance remains pending; optional latent-upscale artifacts remain a separate issue. +- Put the patch summary and v5-to-v6 restart instructions at the beginning of both language READMEs. +- Replace workstation paths and private operational history in public documentation with portable instructions and public validation summaries. Original development records are preserved outside the publication checkout. +- Mark the eighteen real-Core integration cases as dependency skips in standalone CI when ComfyUI Core is absent. They still execute with Core installed; broken Core dependencies remain test failures. Production code is unchanged. +- Release-preparation checks for the affected tests and package metadata: 66 passed with real Core; 48 passed / 18 dependency skips without Core. Both CPU gates had zero CUDA initialization attempts. +- GitHub Actions confirmed 1,676 passed / 21 skipped / no failures in standalone CI. Eighteen skips require real Core; the remaining three reflect existing platform/fixture limits. Prepare the versioned GitHub Release with English/Japanese repair and migration notes; Registry publication remains separate. +- Reconcile PACKAGE_VALIDATION.txt, PACKAGE_INFO_JA.txt and docs/VALIDATION_JA.md with package 3.9.1, the actual CPU/CI scopes, v5 migration and Release status. Retain the original V3.9.0/V2 text under explicit historical sections. Documentation corrections follow the fixed v3.9.1 release tag on main; runtime and official workflow bytes are unchanged. + +Earlier product behavior is described in CHANGELOG.md and the historical Release/tag versions. This log contains public engineering summaries rather than private operational records. + + +## 2026-10-13 — Prepare 3.9.3 versioned downloads + +- Add 3.9.3 and retained 3.9.2 Release, custom-node ZIP and checksum links to the current English/Japanese guides. Update the installation heading to 3.9.3 and retain the implemented settings screenshot and historical validation records. +- Refresh Registry then development manifests from saved bytes. The existing release-metadata/public-surface CPU gate passes 8 tests with zero CUDA initialization attempts; all 25 tracked JSON/ZIP assets retain their bytes. +- Product code, package metadata, public schemas and saved-data contracts are unchanged. GitHub Release/asset and Registry publication are verified separately; no new GPU, long-form performance or physical SSD-write validation is claimed. + + +## 2026-10-15 — Issue #31 Spectrum binding identity candidate + +Added a dedicated, fail-closed Spectrum binding adapter without changing the generic observer, Sampling v6, runtime signature v1, Graph v4, public nodes, workflows or package version. The immutable config fields and audited source/loaded-callable contract identify generation; reviewed runtime history, counters and profile caches do not change persistent identity. Unsupported structures, additional attributes and altered implementations remain non-reusable. Existing unsafe Takes are preserved without migration. CPU and GPU evidence are recorded separately; no commit or publication is authorized. + + +Issue #31 Phase 3 validation: the final standard CPU entry point passed 1,834 tests, 1 skip and 2 existing expected Inspector failures, with zero CUDA initialization attempts. The 57 new Spectrum tests cover all 27 config fields, transient-state/process stability, implementation/extra-state refusal, unaffected v6 contract identity and old unsafe Take protection; the 36 Phase 1 attention tests also pass. Under one audited local provider implementation, both PyTorch+Spectrum and Kitchen+Spectrum persisted new Chunk 1 Takes and reused them from a separate Queue to generate Chunk 2, with unchanged prefix hashes, identical safe contracts and fully decodable 10-second AV output. This is a targeted local GPU gate, not acceptance of other providers, KJ, browser behavior, subjective quality or benchmarks. No commit, push or publication occurred. + + +## 2026-10-20 — External Chunks Reference assignment table candidate + +Only the V3.9 Reference Images helper presentation changes: external Chunks links make the destination count unresolved, expose assignment range 1–16, and label All/Off with that range. Direct counts retain the existing table. The frontend does not resolve arbitrary upstream values or alter native Chunks, saved selectors, Sampling identity or stored Takes. Added 21 JavaScript cases and four public-input CPU cases cover stale widget counts, zero/unrestored links, multiple Samplers, range actions, future assignments and callback stability. Existing CPU coverage passed 310 tests with two known Inspector expected failures; the four added CPU cases and metadata checks also pass. Targeted browser acceptance passed for external INT with stale widget values, live disconnection/restoration, actual KJ Set/Get routing, all-page All/Off actions, saving/reloading and workflow tab switching. The candidate display code was deployed to the chosen runtime and reloaded without restarting its backend. Existing saved Takes and workflows were preserved. No GPU generation, commit, push or public release occurred. + +## 2026-10-20 — Spectrum source/runtime audit portability candidate + +Normalize CRLF to LF in Spectrum provider-file and module-closure source checks. Add fixed Python 3.12 profiles generated from the same hash-verified 44-file and upstream 0.2.29 55-file sources. Retain the original Python 3.13 profile bytes and valid identity descriptors. Unknown code, file sets, closures and runtime extensions remain non-reusable; unsupported Python runtimes are not promoted. Old unsafe/v5 Takes remain readable without conversion, and public nodes, Sampling v6, Graph v4 and official workflows are unchanged. + +The full standard CPU/JS gate passed 2,096 tests with two skips and two existing Inspector expected failures; compile and offline runtime verification passed with twelve public nodes. Targeted Spectrum/Attention coverage passed 169 tests with one legacy-profile skip. Python 3.12 component coverage passed 112 tests with one legacy-profile skip, checking all 27 configuration fields, code/state refusal and newline parity; 21 independent older-Core/Attention cases were outside that component gate. Separate-process storage checks passed 16 PyTorch/Kitchen cases on Python 3.13 and eight Spectrum/default-model cases on Python 3.12. Actual Take read-only JavaScript agreed with all 24 storage results, and all six previously valid complete Sampling Contracts remained identical. No CUDA initialization, new GPU generation, browser acceptance, commit, push or publication is claimed. The older Core/Kitchen observation limitation remains separate from Spectrum recognition. + +## 2026-10-20 — 3.9.10 publication preparation + +Prepare the tested Spectrum audit portability and Reference assignment display +repairs as the next 3.9 patch. Use SemVer 3.9.10 for GitHub and Registry while +reserving 4.0 for the major update. Preserve existing 3.9.9 and historical tags, +assets and official workflow bytes. Publish CPU/browser scope explicitly without +claiming new GPU acceptance. Version metadata, release workflows and package +manifests identify the same candidate; public asset and CI checks follow upload. diff --git a/__init__.py b/__init__.py new file mode 100644 index 0000000..f0382e9 --- /dev/null +++ b/__init__.py @@ -0,0 +1,32 @@ +"""H3 Continuum — native MiniMax H3 chunk continuation.""" + +from __future__ import annotations + +import logging + +WEB_DIRECTORY = "./web" + +# ComfyUI loads custom-node folders as packages. Standalone test collection may +# import this file as a top-level module; keep that path inert because ComfyUI is +# intentionally not a unit-test dependency. +if __package__: + from .nodes import NODE_CLASS_MAPPINGS, NODE_DISPLAY_NAME_MAPPINGS + from .temporal import run_temporal_self_test + from .version import PACKAGE_VERSION + from .storage_api import register_routes + + register_routes() + + try: + run_temporal_self_test() + except Exception as exc: + raise RuntimeError(f"H3 Continuum Join self-test failed: {exc}") from exc + logging.getLogger("h3_continuum_join").info( + "H3 Continuum %s loaded (V2 integrated sampler + hidden legacy workflow compatibility)", + PACKAGE_VERSION, + ) +else: # pragma: no cover + NODE_CLASS_MAPPINGS = {} + NODE_DISPLAY_NAME_MAPPINGS = {} + +__all__ = ["NODE_CLASS_MAPPINGS", "NODE_DISPLAY_NAME_MAPPINGS", "WEB_DIRECTORY"] diff --git a/audio_compat.py b/audio_compat.py new file mode 100644 index 0000000..39489a1 --- /dev/null +++ b/audio_compat.py @@ -0,0 +1,29 @@ +"""Use Core audio DSP; retain the existing adapter for older Core versions.""" + +from __future__ import annotations + +import importlib + + +def resample_audio(waveform, source_rate: int, target_rate: int): + if source_rate == target_rate: + return waveform + try: + core_audio = importlib.import_module("comfy.audio") + except ModuleNotFoundError as exc: + if exc.name not in {"comfy", "comfy.audio"}: + raise + core_audio = None + core_resample = getattr(core_audio, "resample", None) + if core_resample is not None: + # Errors from an available Core API are real errors, not API absence. + return core_resample(waveform, source_rate, target_rate) + try: + torchaudio = importlib.import_module("torchaudio") + except ModuleNotFoundError as exc: + if exc.name != "torchaudio": + raise + raise RuntimeError( + "Audio resampling requires comfy.audio.resample or legacy torchaudio" + ) from exc + return torchaudio.functional.resample(waveform, source_rate, target_rate) diff --git a/branch_provenance.py b/branch_provenance.py new file mode 100644 index 0000000..e5b8806 --- /dev/null +++ b/branch_provenance.py @@ -0,0 +1,222 @@ +"""Pure Branch Provenance Contract v1 helpers. + +The contract describes immutable physical-group revisions independently from +the sampling contract that produced their tensors. A display label such as +``Take B`` is deliberately not persisted as identity. +""" + +from __future__ import annotations + +import hashlib +import json +from collections.abc import Iterable, Mapping +from dataclasses import dataclass +from typing import Any + + +BRANCH_PROVENANCE_VERSION = 1 +TAKE_ACTION_AUTOMATIC = "Automatic" +TAKE_ACTION_USE = "Use This Take" +TAKE_ACTION_CONTINUE = "Continue From Here" +TAKE_ACTION_OPTIONS = ( + TAKE_ACTION_AUTOMATIC, + TAKE_ACTION_USE, + TAKE_ACTION_CONTINUE, +) + + +class BranchProvenanceError(ValueError): + """Raised when persisted branch ancestry is ambiguous or inconsistent.""" + + +def _canonical(value: Any) -> str: + return json.dumps(value, sort_keys=True, separators=(",", ":"), ensure_ascii=True) + + +def _hash(value: Any) -> str: + return hashlib.sha256(_canonical(value).encode("utf-8")).hexdigest() + + +@dataclass(frozen=True) +class PhysicalGroup: + start: int + end: int + physical_group: int + + def as_metadata(self) -> dict[str, int]: + return { + "start": self.start, + "end": self.end, + "physical_group": self.physical_group, + } + + +def physical_groups(*, chunks: int, terminal_merge_enabled: bool) -> tuple[PhysicalGroup, ...]: + if isinstance(chunks, bool) or not isinstance(chunks, int) or chunks < 1: + raise BranchProvenanceError("chunks must be a positive integer") + if type(terminal_merge_enabled) is not bool: + raise BranchProvenanceError("terminal_merge_enabled must be a boolean") + if terminal_merge_enabled and chunks < 2: + raise BranchProvenanceError("Terminal Merge requires at least two chunks") + terminal_start = chunks - 1 if terminal_merge_enabled else None + groups: list[PhysicalGroup] = [] + chunk = 1 + while chunk <= chunks: + if terminal_start is not None and chunk == terminal_start: + groups.append(PhysicalGroup(chunk, chunks, chunk)) + break + groups.append(PhysicalGroup(chunk, chunk, chunk)) + chunk += 1 + return tuple(groups) + + +def _record_identity(record: Mapping[str, Any]) -> dict[str, Any]: + entry = record.get("entry") or {} + return { + "sequence_index": int(record.get("sequence_index", -1)), + "storage_revision_id": str(record.get("storage_revision_id", "")), + "filename": str(record.get("filename", "")), + "file_size": int(record.get("file_size", -1)), + "file_sha256": str(record.get("file_sha256", "")), + "seed": int(record.get("seed", entry.get("seed", -1))), + "prompt_hash": str(record.get("prompt_hash", entry.get("prompt_hash", ""))), + } + + +def make_group_revision( + *, + parent_revision_id: str | None, + group: PhysicalGroup, + variation_nonce: int, + created_utc: str, + storage_revision_id: str, + generation_contract_sha256: str, + lineage_sha256: str, + records: Iterable[Mapping[str, Any]], + branch_cut: Mapping[str, Any] | None = None, +) -> dict[str, Any]: + if not isinstance(group, PhysicalGroup): + raise BranchProvenanceError("group must be a PhysicalGroup") + nonce = int(variation_nonce) + if nonce < 0 or nonce > 0xFFFFFFFF: + raise BranchProvenanceError("variation_nonce is outside uint32") + record_values = [_record_identity(record) for record in records] + expected_positions = list(range(group.start - 1, group.end)) + if [record["sequence_index"] for record in record_values] != expected_positions: + raise BranchProvenanceError("physical group records are incomplete or non-atomic") + if not all(record["storage_revision_id"] for record in record_values): + raise BranchProvenanceError("physical group record source is missing") + seed_identity = _hash( + [ + { + "sequence_index": record["sequence_index"], + "seed": record["seed"], + "prompt_hash": record["prompt_hash"], + } + for record in record_values + ] + ) + identity = { + "branch_provenance_version": BRANCH_PROVENANCE_VERSION, + "parent_revision_id": None if parent_revision_id is None else str(parent_revision_id), + "group": group.as_metadata(), + "variation_nonce": nonce, + "seed_identity": seed_identity, + "storage_revision_id": str(storage_revision_id), + "generation_contract_sha256": str(generation_contract_sha256), + "lineage_sha256": str(lineage_sha256), + "records": record_values, + } + revision_id = _hash(identity)[:24] + result = { + "revision_id": revision_id, + **identity, + "created_utc": str(created_utc), + "revision_order": f"{created_utc}|{revision_id}", + } + if branch_cut is not None: + result["branch_cut"] = { + "selected_revision_id": str(branch_cut.get("selected_revision_id", "")), + "after_physical_group": int(branch_cut.get("after_physical_group", 0)), + } + return result + + +def validate_group_revision(value: Mapping[str, Any]) -> dict[str, Any]: + if not isinstance(value, Mapping): + raise BranchProvenanceError("group revision must be an object") + if int(value.get("branch_provenance_version", -1)) != BRANCH_PROVENANCE_VERSION: + raise BranchProvenanceError("group revision provenance version is unsupported") + group_value = value.get("group") + if not isinstance(group_value, Mapping): + raise BranchProvenanceError("group revision physical group is missing") + group = PhysicalGroup( + start=int(group_value.get("start", 0)), + end=int(group_value.get("end", 0)), + physical_group=int(group_value.get("physical_group", 0)), + ) + if group.start < 1 or group.end < group.start or group.physical_group != group.start: + raise BranchProvenanceError("group revision physical group is invalid") + rebuilt = make_group_revision( + parent_revision_id=value.get("parent_revision_id"), + group=group, + variation_nonce=int(value.get("variation_nonce", -1)), + created_utc=str(value.get("created_utc", "")), + storage_revision_id=str(value.get("storage_revision_id", "")), + generation_contract_sha256=str(value.get("generation_contract_sha256", "")), + lineage_sha256=str(value.get("lineage_sha256", "")), + records=value.get("records") or [], + branch_cut=value.get("branch_cut"), + ) + if str(value.get("revision_id", "")) != rebuilt["revision_id"]: + raise BranchProvenanceError("group revision immutable identity is invalid") + if str(value.get("seed_identity", "")) != rebuilt["seed_identity"]: + raise BranchProvenanceError("group revision seed identity is invalid") + rebuilt["revision_order"] = str( + value.get("revision_order", rebuilt["revision_order"]) + ) + return rebuilt + + +def resolve_chain( + catalog: Mapping[str, Mapping[str, Any]], + head_revision_id: str, +) -> list[dict[str, Any]]: + head = str(head_revision_id) + if not head: + raise BranchProvenanceError("selected revision ID is empty") + reverse: list[dict[str, Any]] = [] + seen: set[str] = set() + current: str | None = head + while current is not None: + if current in seen: + raise BranchProvenanceError("group revision ancestry contains a cycle") + seen.add(current) + raw = catalog.get(current) + if raw is None: + raise BranchProvenanceError( + f"group revision parent is missing: {current}" + ) + revision = validate_group_revision(raw) + reverse.append(revision) + parent = revision.get("parent_revision_id") + current = None if parent is None else str(parent) + chain = list(reversed(reverse)) + expected_start = 1 + lineage = str(chain[0].get("lineage_sha256", "")) + for revision in chain: + group = revision["group"] + if int(group["start"]) != expected_start: + raise BranchProvenanceError("group revision ancestry is not a contiguous prefix") + if str(revision.get("lineage_sha256", "")) != lineage: + raise BranchProvenanceError("group revision ancestry crosses sampling lineages") + expected_start = int(group["end"]) + 1 + return chain + + +def active_revision_map(chain: Iterable[Mapping[str, Any]]) -> dict[str, str]: + active: dict[str, str] = {} + for revision in chain: + validated = validate_group_revision(revision) + active[str(validated["group"]["physical_group"])] = validated["revision_id"] + return active diff --git a/compatibility.py b/compatibility.py new file mode 100644 index 0000000..40fe39a --- /dev/null +++ b/compatibility.py @@ -0,0 +1,291 @@ +"""ComfyUI native-H3 and accelerator compatibility checks.""" + +from __future__ import annotations + +import inspect +from typing import Any, Iterable + + +class CompatibilityError(RuntimeError): + """Raised when a MODEL is not the supported native MiniMax H3 runtime.""" + + +def _missing_callable_parameters( + callable_obj: Any, + *, + positional: Iterable[str] = (), + keywords: Iterable[str] = (), +) -> list[str]: + """Return parameters that a callable can no longer accept. + + Compatibility wrappers are allowed to replace a concrete signature with + ``*args``/``**kwargs``. Sol-Attn's MiniMax-H3 Morton integration does this + for ``PackedLayout.__init__`` while transparently forwarding the original + arguments. Treating the wrapped signature as an API removal produces a + false incompatibility even though construction remains valid. + + Explicit parameters are preferred when available. ``*args`` satisfies + missing positional parameters and ``**kwargs`` satisfies missing keyword + parameters. If ``inspect.signature`` itself is unavailable, the caller + still receives the exception and may fail closed. + """ + + params = inspect.signature(callable_obj).parameters + has_var_positional = any( + parameter.kind is inspect.Parameter.VAR_POSITIONAL + for parameter in params.values() + ) + has_var_keyword = any( + parameter.kind is inspect.Parameter.VAR_KEYWORD + for parameter in params.values() + ) + + missing: list[str] = [] + for name in positional: + if name not in params and not has_var_positional: + missing.append(name) + for name in keywords: + if name not in params and not has_var_keyword: + missing.append(name) + return missing + + +def check_comfy_h3_runtime() -> list[str]: + """Return human-readable incompatibilities with ComfyUI's native H3 API. + + The check deliberately validates public/runtime contracts only. It does not + import optional accelerators and therefore cannot make their installation a + hard dependency. + """ + + issues: list[str] = [] + try: + import comfy.ldm.minimax.model as h3_model + except Exception as exc: # pragma: no cover - requires a broken Comfy install + return [f"cannot import native MiniMax H3: {exc}"] + + layout_cls = getattr(h3_model, "PackedLayout", None) + model_cls = getattr(h3_model, "MiniMaxH3Model", None) + if layout_cls is None: + issues.append("PackedLayout is missing") + else: + try: + missing = _missing_callable_parameters( + layout_cls.__init__, + positional=( + "text_len", + "latent_t", + "latent_h", + "latent_w", + "audio_t", + ), + keywords=( + "keyframes", + "refs", + ), + ) + for name in missing: + issues.append(f"PackedLayout no longer accepts '{name}'") + except (TypeError, ValueError) as exc: + issues.append(f"cannot inspect PackedLayout: {exc}") + + if model_cls is None: + issues.append("MiniMaxH3Model is missing") + else: + for name in ( + "forward", + "rope_freqs", + "_cond_video_rows", + "_cond_audio_rows", + ): + if not hasattr(model_cls, name): + issues.append(f"MiniMaxH3Model.{name} is missing") + + if tuple(getattr(h3_model, "FRAME_PER_TOKEN", ())) != (1, 4, 4, 4, 4): + issues.append("MiniMax H3 temporal compression grid changed") + + try: + import comfy.model_base as model_base + + base_cls = getattr(model_base, "MiniMaxH3", None) + if base_cls is None or not hasattr(base_cls, "extra_conds"): + issues.append("native MiniMaxH3 BaseModel.extra_conds is missing") + except Exception as exc: + issues.append(f"cannot inspect MiniMaxH3 BaseModel: {exc}") + + try: + from comfy.patcher_extension import WrappersMP + + for name in ("APPLY_MODEL", "DIFFUSION_MODEL"): + if not hasattr(WrappersMP, name): + issues.append(f"WrappersMP.{name} is missing") + except Exception as exc: + issues.append(f"cannot import model wrapper API: {exc}") + + try: + import comfy.nested_tensor + + if not hasattr(comfy.nested_tensor, "NestedTensor"): + issues.append("comfy.nested_tensor.NestedTensor is missing") + except Exception as exc: + issues.append(f"cannot import NestedTensor API: {exc}") + return issues + + + +def run_native_layout_self_test() -> str: + """Exercise the installed native H3 PackedLayout without loading a model. + + This catches the most dangerous class of upstream change: a layout that can + still be constructed but no longer maps reference/target rows the way the + Continuum adapter expects. + """ + + import torch + import comfy.ldm.minimax.model as h3_model + + from .constants import ( + MARK_AUDIO_CONTEXT, + MARK_AUDIO_END_FRAME, + MARK_AUDIO_OVERHANG, + MARK_CONTEXT_FRAMES, + MARK_VIDEO_CONTEXT, + ) + from .layout_adapter import normalize_condition_latents, patch_layout_in_place + + video = torch.zeros(1, 24, 2, 2, 2) # five-frame context + audio = torch.zeros(1, 32, 2, 8) + ref = { + "kind": "video_audio", + "latent_t": 2, + "latent_h": 2, + "latent_w": 2, + "ref_audio_t": 8, + "latent": video, + "audio_latent": audio, + MARK_VIDEO_CONTEXT: True, + MARK_CONTEXT_FRAMES: 5, + MARK_AUDIO_CONTEXT: True, + MARK_AUDIO_END_FRAME: 5.0, + MARK_AUDIO_OVERHANG: -1.0 / 3.0, + } + layout = h3_model.PackedLayout( + 3, 7, 2, 2, 10, keyframes=None, refs=[ref] + ) + position_id = id(layout.position_ids) + payload = {"layout": layout, "keyframes": [], "refs": [ref]} + normalize_condition_latents(payload) + result = patch_layout_in_place(payload, strict=True, debug=False) + if id(layout.position_ids) != position_id: + raise CompatibilityError("layout self-test replaced position_ids") + ref_video = next((a, b) for a, b, kind in layout.segments if kind == "ref_img") + target_video = next((a, b) for a, b, kind in layout.segments if kind == "video") + ra, rb = ref_video + va, _vb = target_video + if not torch.equal(layout.position_ids[ra:rb], layout.position_ids[va : va + (rb - ra)]): + raise CompatibilityError("layout self-test could not align video context") + audio_refs = [ + { + "kind": "audio", + "ref_audio_t": length, + "audio_latent": torch.zeros(1, 32, 2, length), + } + for length in (2, 3, 4) + ] + audio_layout = h3_model.PackedLayout( + 3, 7, 2, 2, 10, keyframes=None, refs=audio_refs + ) + audio_segments = [ + stop - start + for start, stop, kind in audio_layout.segments + if kind == "ref_audio" + ] + packed_audio_rows = torch.cat( + [h3_model.pack_audio(item["audio_latent"]) for item in audio_refs], dim=0 + ).shape[0] + if audio_segments != [4, 6, 8] or sum(audio_segments) != packed_audio_rows: + raise CompatibilityError( + "layout self-test found a multi-Reference-Audio packed-row mismatch" + ) + return ( + f"native PackedLayout self-test passed; rows={layout.position_ids.shape[0]}, " + f"position_ids_id={result['position_ids_id']}, " + f"reference_audio_rows={packed_audio_rows}" + ) + +def ensure_native_h3_base_model(base_model: Any) -> None: + """Validate a ComfyUI BaseModel instance before adding our wrapper.""" + + inner = getattr(base_model, "diffusion_model", None) + if inner is None: + raise CompatibilityError("MODEL does not expose diffusion_model") + missing = [ + name + for name in ( + "blocks", + "final_layer", + "patch_size", + "latents_dim", + "audio_latents_dim", + "rope_freqs", + ) + if not hasattr(inner, name) + ] + if missing: + actual = f"{type(inner).__module__}.{type(inner).__name__}" + raise CompatibilityError( + f"MiniMax H3 model contract is unavailable on {actual}; missing " + + ", ".join(missing) + ) + + +def _wrapper_keys(model: Any) -> tuple[str, ...]: + """Collect ModelPatcher wrapper keys without depending on enum values.""" + + wrappers = getattr(model, "wrappers", {}) or {} + found: list[str] = [] + if not isinstance(wrappers, dict): + return () + for group in wrappers.values(): + if not isinstance(group, dict): + continue + found.extend(str(key) for key in group.keys()) + return tuple(found) + + +def _contains_any(values: Iterable[str], *needles: str) -> bool: + text = " ".join(values).lower() + return any(needle.lower() in text for needle in needles) + + +def accelerator_summary(model: Any) -> str: + """Best-effort, non-authoritative summary of attached accelerators. + + Optional custom nodes expose different public markers across releases. The + summary therefore reports only what is observable and never gates execution. + """ + + options = getattr(model, "model_options", {}) or {} + transformer = options.get("transformer_options", {}) or {} + wrapper_keys = _wrapper_keys(model) + parts: list[str] = [] + + if _contains_any(wrapper_keys, "spectrum"): + parts.append("Spectrum wrapper detected") + + sol_marker = transformer.get("sol_compose") is not None or _contains_any( + wrapper_keys, "sol_attn", "sol-attn", "solattn" + ) + if sol_marker: + parts.append("Sol-Attn marker detected") + + if transformer.get("optimized_attention_override") is not None: + if sol_marker: + parts.append("chained attention override present (Sage/other fallback possible)") + else: + parts.append("attention override present (Sage/other)") + + if options.get("h3_continuum_join"): + parts.append("Continuum APPLY_MODEL wrapper installed") + + return "; ".join(parts) if parts else "accelerator markers not detected (informational only)" diff --git a/conditioning.py b/conditioning.py new file mode 100644 index 0000000..b617609 --- /dev/null +++ b/conditioning.py @@ -0,0 +1,105 @@ +from __future__ import annotations + +from typing import Any + + +CONDITIONING_MODE_T2VA = "t2va" +CONDITIONING_MODE_I2VA = "i2va" +CONDITIONING_MODE_LAST_ONLY = "last_only" +CONDITIONING_MODE_FL2VA = "fl2va" +CONDITIONING_MODE_REFERENCE = "reference" + +CONDITIONING_MODES = frozenset( + { + CONDITIONING_MODE_T2VA, + CONDITIONING_MODE_I2VA, + CONDITIONING_MODE_LAST_ONLY, + CONDITIONING_MODE_FL2VA, + CONDITIONING_MODE_REFERENCE, + } +) + +_MODE_LABELS = { + CONDITIONING_MODE_T2VA: "T2VA + Continuation", + CONDITIONING_MODE_I2VA: "I2VA + Continuation", + CONDITIONING_MODE_LAST_ONLY: "Last Frame + Continuation", + CONDITIONING_MODE_FL2VA: "FL2VA + Continuation", + CONDITIONING_MODE_REFERENCE: "Reference + Continuation", +} + + +def conditioning_mode_from_presence( + *, has_first: bool, has_last: bool, has_reference: bool +) -> str: + if has_reference: + return CONDITIONING_MODE_REFERENCE + if has_first and has_last: + return CONDITIONING_MODE_FL2VA + if has_first: + return CONDITIONING_MODE_I2VA + if has_last: + return CONDITIONING_MODE_LAST_ONLY + return CONDITIONING_MODE_T2VA + + +def detect_conditioning_mode( + *, first_frame: Any, last_frame: Any, reference_assets: Any +) -> str: + return conditioning_mode_from_presence( + has_first=first_frame is not None, + has_last=last_frame is not None, + has_reference=reference_assets is not None, + ) + + +def conditioning_mode_label(mode: str) -> str: + try: + return _MODE_LABELS[str(mode)] + except KeyError as exc: + raise ValueError(f"unknown conditioning mode: {mode!r}") from exc + + +def conditioning_display_label( + *, + has_first: bool, + has_last: bool, + has_reference: bool, +) -> str: + """Return a user-facing label without changing the runtime mode contract.""" + if has_reference: + if has_first and has_last: + return "Hybrid FL2VA + Reference + Continuation" + if has_first: + return "Hybrid I2VA + Reference + Continuation" + if has_last: + return "Hybrid L2VA + Reference + Continuation" + return "Reference + Continuation" + mode = conditioning_mode_from_presence( + has_first=has_first, + has_last=has_last, + has_reference=False, + ) + return conditioning_mode_label(mode) + + +def conditioning_mode_display_label( + mode: str, + *, + has_first: bool = False, + has_last: bool = False, + has_reference: bool = False, +) -> str: + """Legacy display helper retained for existing callers and tests.""" + if has_first or has_last or has_reference: + return conditioning_display_label( + has_first=has_first, + has_last=has_last, + has_reference=has_reference, + ) + return conditioning_mode_label(mode) + + +def conditioning_mode_uses_video_vae(mode: str) -> bool: + if mode not in CONDITIONING_MODES: + raise ValueError(f"unknown conditioning mode: {mode!r}") + return mode != CONDITIONING_MODE_T2VA diff --git a/constants.py b/constants.py new file mode 100644 index 0000000..8abd6f8 --- /dev/null +++ b/constants.py @@ -0,0 +1,123 @@ +"""Constants shared by H3 Continuum Join.""" + +from __future__ import annotations + +FPS = 24.0 +AUDIO_SAMPLE_RATE = 32_000 +AUDIO_LATENT_FPS = 40.0 +FRAME_RESCALE = AUDIO_LATENT_FPS / FPS +FRAME_PER_TOKEN = (1, 4, 4, 4, 4) + +CHUNK_SECONDS_DEFAULT = 5.0 +CHUNK_SECONDS_MIN = 4.0 +CHUNK_SECONDS_MAX = 30.0 +CHUNK_SECONDS_STEP = 0.1 +CHUNK_SECONDS_TOOLTIP = ( + "Target duration shared by every chunk. 5–15 seconds is the recommended " + "and validated range. Values above 15 seconds are supported, but VRAM " + "use and processing time can increase substantially, especially at high " + "resolution." +) + +STATE_MAGIC = "H3_CONTINUUM_STATE" +PLAN_MAGIC = "H3_CONTINUUM_PLAN" +SESSION_MAGIC = "H3_CONTINUUM_SESSION" +PROMPT_PLAN_MAGIC = "H3_CONTINUUM_PROMPT_PLAN" + +MARK_SCHEMA = "_h3cj_schema" +MARK_VIDEO_SLOT = "_h3cj_video_slot" # legacy/internal-slot compatibility +MARK_PIXEL_START = "_h3cj_pixel_start" +MARK_VIDEO_CONTEXT = "_h3cj_video_context" +MARK_CONTEXT_FRAMES = "_h3cj_context_frames" +MARK_AUDIO_CONTEXT = "_h3cj_audio_context" +MARK_AUDIO_END_FRAME = "_h3cj_audio_end_frame" +MARK_AUDIO_OVERHANG = "_h3cj_audio_overhang" +CONTINUUM_REFERENCE_METADATA_KEY = "_h3_continuum" +CONTINUUM_INTEROP_KEY = "h3_continuum" +CONTINUUM_INTEROP_API = 1 +CONTINUUM_ACTUAL_PREFIX_STEPS = 2 + +MODEL_WRAPPER_KEY = "h3_continuum_join.apply_model.v1" +V2_RUNTIME_KEY = "h3_continuum_v2" +LAYOUT_ORIGINAL_TIME_ATTR = "_h3cj_original_time_v1" +LAYOUT_SIGNATURE_ATTR = "_h3cj_patch_signature_v1" +LAYOUT_DEVICE_CACHE_ATTR = "_h3cj_device_cache_v1" + +CONTINUITY_OPTIONS = ( + "Balanced — 22 frames", + "Fast — 5 frames", + "Strong — 39 frames (Experimental)", +) +CONTINUITY_FRAMES = { + CONTINUITY_OPTIONS[0]: 22, + CONTINUITY_OPTIONS[1]: 5, + CONTINUITY_OPTIONS[2]: 39, +} +DEFAULT_STATE_CAPACITY_FRAMES = 39 + +V2_CONTINUITY_AUTO = "Auto — conservative" +V2_CONTINUITY_OPTIONS = (V2_CONTINUITY_AUTO,) + CONTINUITY_OPTIONS + +# Older saved workflows and frontend help text used ASCII hyphens in these +# labels. Keep the public labels canonical while accepting those persisted +# aliases when a saved Run Storage session is resumed. +_CONTINUITY_ALIASES = { + "Auto - conservative": V2_CONTINUITY_AUTO, + "Balanced - 22 frames": CONTINUITY_OPTIONS[0], + "Fast - 5 frames": CONTINUITY_OPTIONS[1], + "Strong - 39 frames (Experimental)": CONTINUITY_OPTIONS[2], +} + + +def normalize_continuity_mode(value: str) -> str: + """Return the canonical continuity label for current and legacy saves.""" + + raw = str(value).strip() + return _CONTINUITY_ALIASES.get(raw, raw) + + +def continuity_storage_mode(value: str) -> str: + """Return the stable ASCII label used in persisted contracts. + + Persisted V3.8 runs used ASCII hyphens. Keeping that storage spelling + makes old manifests compatible while the runtime continues to accept and + expose the public em-dash labels. + """ + + return normalize_continuity_mode(value).replace("—", "-") + +PROMPT_MODE_FIXED = "Fixed — one prompt" +PROMPT_MODE_LIST = "List — split with ---" +PROMPT_MODE_TIMELINE = "Timeline — [0-5s] sections" +PROMPT_MODE_OPTIONS = (PROMPT_MODE_FIXED, PROMPT_MODE_LIST, PROMPT_MODE_TIMELINE) + +PROMPT_FORMAT_AUTO = "Auto" +PROMPT_FORMAT_FIXED = "Fixed" +PROMPT_FORMAT_LIST = "List" +PROMPT_FORMAT_TIMELINE = "Timeline" +PROMPT_FORMAT_OPTIONS = ( + PROMPT_FORMAT_AUTO, + PROMPT_FORMAT_FIXED, + PROMPT_FORMAT_LIST, + PROMPT_FORMAT_TIMELINE, +) + +DIAGNOSTICS_OFF = "Off" +DIAGNOSTICS_BASIC = "Basic" +DIAGNOSTICS_FULL = "Detailed Report" +DIAGNOSTICS_LEGACY_FULL = "Full" +DIAGNOSTICS_OPTIONS = (DIAGNOSTICS_BASIC, DIAGNOSTICS_FULL, DIAGNOSTICS_OFF) +DIAGNOSTICS_ACCEPTED_OPTIONS = DIAGNOSTICS_OPTIONS + (DIAGNOSTICS_LEGACY_FULL,) + + +def normalize_diagnostics_mode(value: str) -> str: + return DIAGNOSTICS_FULL if value == DIAGNOSTICS_LEGACY_FULL else value + +SEAM_CORRECTION_AUTO = "Auto" +SEAM_CORRECTION_OFF = "Off" +SEAM_CORRECTION_BASIC = "Basic" +SEAM_CORRECTION_OPTIONS = ( + SEAM_CORRECTION_AUTO, + SEAM_CORRECTION_OFF, + SEAM_CORRECTION_BASIC, +) diff --git a/continuation.py b/continuation.py new file mode 100644 index 0000000..a672664 --- /dev/null +++ b/continuation.py @@ -0,0 +1,169 @@ +"""Reusable native-H3 continuation preparation independent of ComfyUI UI nodes.""" + +from __future__ import annotations + +from typing import Any + +import torch + +from .constants import ( + CONTINUUM_INTEROP_API, + CONTINUUM_REFERENCE_METADATA_KEY, + MARK_AUDIO_CONTEXT, + MARK_AUDIO_END_FRAME, + MARK_AUDIO_OVERHANG, + MARK_CONTEXT_FRAMES, + MARK_SCHEMA, + MARK_VIDEO_CONTEXT, +) +from .state import extract_av_streams +from .version import STATE_SCHEMA_VERSION + +POLICY_REPLACE = "Use Continuum context; keep visual identity tokens (Recommended)" +POLICY_ERROR = "Error if a first-frame keyframe exists" +POLICY_KEEP = "Keep first-frame keyframe too (Advanced)" +FIRST_FRAME_POLICIES = (POLICY_REPLACE, POLICY_ERROR, POLICY_KEEP) + + +def clone_conditioning(conditioning): + """Clone the outer conditioning list and metadata dictionaries.""" + output = [] + for index, item in enumerate(conditioning): + if not isinstance(item, (tuple, list)) or len(item) != 2: + raise ValueError(f"conditioning entry {index} is not [tensor, metadata]") + if not isinstance(item[1], dict): + raise ValueError(f"conditioning entry {index} metadata is not a dictionary") + output.append([item[0], dict(item[1])]) + if not output: + raise ValueError("conditioning is empty") + return output + + +def conditioning_diagnostics(conditioning) -> dict[str, int]: + """Return non-authoritative cues used in the Join report.""" + first_keyframes = 0 + visual_tokens = 0 + for item in conditioning: + if not isinstance(item, (tuple, list)) or len(item) != 2: + continue + metadata = item[1] if isinstance(item[1], dict) else {} + for keyframe in metadata.get("minimax_keyframes") or (): + if int(keyframe.get("resolved_frame_index", 0)) == 0: + first_keyframes += 1 + tags = metadata.get("minimax_token_tags") + if torch.is_tensor(tags): + visual_tokens += int((tags == 0).sum().item()) + return {"first_keyframes": first_keyframes, "visual_tokens": visual_tokens} + + +def prepare_conditioning( + conditioning, + *, + video_context: torch.Tensor, + audio_context: torch.Tensor | None, + audio_grid_offset: float, + context_frames: int, + new_frame_count: int, + first_frame_policy: str, + preserve_last_frame: bool, + source_frame_count: int | None = None, +): + """Add one native H3 video/video-audio reference as timeline context.""" + if first_frame_policy not in FIRST_FRAME_POLICIES: + raise ValueError(f"unknown first-frame policy: {first_frame_policy!r}") + if not torch.is_tensor(video_context) or video_context.ndim != 5: + raise ValueError("video_context must be [1,24,T,H,W]") + if tuple(video_context.shape[:2]) != (1, 24): + raise ValueError("video_context must have shape [1,24,T,H,W]") + if audio_context is not None: + if not torch.is_tensor(audio_context) or audio_context.ndim != 4: + raise ValueError("audio_context must be [1,32,2,T]") + if tuple(audio_context.shape[:3]) != (1, 32, 2): + raise ValueError("audio_context must have shape [1,32,2,T]") + + context_t = int(video_context.shape[2]) + context_ref: dict[str, Any] = { + "kind": "video_audio" if audio_context is not None else "video", + "latent_t": context_t, + "latent_h": int(video_context.shape[-2]), + "latent_w": int(video_context.shape[-1]), + "ref_audio_t": int(audio_context.shape[-1]) if audio_context is not None else 0, + "latent": video_context.contiguous(), + MARK_SCHEMA: STATE_SCHEMA_VERSION, + MARK_VIDEO_CONTEXT: True, + MARK_CONTEXT_FRAMES: int(context_frames), + CONTINUUM_REFERENCE_METADATA_KEY: { + "api": CONTINUUM_INTEROP_API, + "role": "video_context", + "audio_role": "audio_context" if audio_context is not None else None, + }, + } + if audio_context is not None: + context_ref.update( + { + "audio_latent": audio_context.contiguous(), + MARK_AUDIO_CONTEXT: True, + MARK_AUDIO_END_FRAME: float(context_frames), + MARK_AUDIO_OVERHANG: float(audio_grid_offset), + } + ) + + output = clone_conditioning(conditioning) + for _, metadata in output: + old_frame_count = metadata.get("minimax_frame_count") + if old_frame_count is None: + old_frame_count = source_frame_count + old_keyframes = [dict(item) for item in (metadata.get("minimax_keyframes") or [])] + kept = [] + for keyframe in old_keyframes: + frame_index = int(keyframe.get("resolved_frame_index", 0)) + if frame_index == 0: + if first_frame_policy == POLICY_ERROR: + raise ValueError( + "conditioning already has a first-frame keyframe. Disconnect the image " + "for continuation, or use the recommended identity-token policy." + ) + if first_frame_policy == POLICY_KEEP: + kept.append(keyframe) + continue + is_last = old_frame_count is not None and frame_index == int(old_frame_count) - 1 + if is_last and preserve_last_frame: + keyframe["resolved_frame_index"] = int(new_frame_count) - 1 + kept.append(keyframe) + elif is_last: + continue + else: + # Core H3 accepts arbitrary guide positions. Preserve them and + # let the native keyframe/layout path decide their semantics. + kept.append(keyframe) + if kept: + metadata["minimax_keyframes"] = kept + else: + metadata.pop("minimax_keyframes", None) + metadata["minimax_frame_count"] = int(new_frame_count) + refs = [ + dict(item) + for item in (metadata.get("minimax_refs") or []) + if not (item.get(MARK_VIDEO_CONTEXT) or item.get(MARK_AUDIO_CONTEXT)) + ] + refs.append(dict(context_ref)) + metadata["minimax_refs"] = refs + return output + + +def new_h3_latent(template: dict[str, Any], *, video_t: int, audio_t: int): + """Create an empty native H3 AV latent matching the template geometry/dtype.""" + try: + import comfy.model_management + import comfy.nested_tensor + except Exception as exc: # pragma: no cover + raise RuntimeError(f"ComfyUI H3 latent helpers unavailable: {exc}") from exc + video, audio = extract_av_streams(template) + device = comfy.model_management.intermediate_device() + new_video = torch.zeros( + (1, 24, int(video_t), int(video.shape[-2]), int(video.shape[-1])), + device=device, + dtype=video.dtype, + ) + new_audio = torch.zeros((1, 32, 2, int(audio_t)), device=device, dtype=audio.dtype) + return {"samples": comfy.nested_tensor.NestedTensor((new_video, new_audio))} diff --git a/decode_cache/__init__.py b/decode_cache/__init__.py new file mode 100644 index 0000000..bd52b55 --- /dev/null +++ b/decode_cache/__init__.py @@ -0,0 +1,2 @@ +"""H3 Decode Cache Helper implementation; imports do not load ComfyUI models.""" +__version__ = "3.8.3" diff --git a/decode_cache/identity.py b/decode_cache/identity.py new file mode 100644 index 0000000..226a0e8 --- /dev/null +++ b/decode_cache/identity.py @@ -0,0 +1,212 @@ +"""Content keys for already-selected decode entries; no model or storage mutation.""" +from __future__ import annotations + +import hashlib +import itertools +import json +import math +import threading +import weakref +from collections.abc import Mapping +from typing import Any, Iterator + +import torch + +CONTRACT = "h3-decode-helper-v1" +BLOCK_BYTES = 8 * 1024 * 1024 + + +class CacheBypass(Exception): + """The native decoder may run, but cache identity is not safely available.""" + + +class ObjectTokens: + """Monotonic process-local identities, without retaining model objects.""" + + def __init__(self): + self._refs = {} + self._next = itertools.count(1) + self._lock = threading.RLock() + + def token(self, obj): + address = id(obj) + with self._lock: + record = self._refs.get(address) + if record is not None and record[0]() is obj: + return record[1] + token = next(self._next) + + def expired(ref, address=address): + with self._lock: + current = self._refs.get(address) + if current is not None and current[0] is ref: + self._refs.pop(address, None) + + try: + ref = weakref.ref(obj, expired) + except TypeError as exc: + raise CacheBypass("object does not support a weak identity") from exc + self._refs[address] = (ref, token) + return token + + +TOKENS = ObjectTokens() + + +def tensor_blocks(tensor: torch.Tensor, block_bytes=BLOCK_BYTES) -> Iterator[memoryview]: + """Yield C-order bytes, including BF16, with bounded contiguous scratch space.""" + if not isinstance(tensor, torch.Tensor) or tensor.layout != torch.strided: + raise CacheBypass("only dense strided tensors are cacheable") + if tensor.device.type != "cpu" or tensor.is_nested or tensor.is_quantized: + raise CacheBypass("non-CPU/nested/quantized payload is delegated without cache") + tensor = tensor.detach() + if tensor.is_conj(): + tensor = tensor.resolve_conj() + if tensor.is_neg(): + tensor = tensor.resolve_neg() + size = tensor.numel() * tensor.element_size() + if size == 0: + return + if size <= block_bytes: + raw = tensor.contiguous().reshape(-1).view(torch.uint8).numpy() + yield memoryview(raw).cast("B") + return + # All earlier dimensions are singleton, preserving logical C-order. + for axis, length in enumerate(tensor.shape): + if length > 1: + per_slice = size // length + step = max(1, block_bytes // max(1, per_slice)) + for start in range(0, length, step): + yield from tensor_blocks(tensor.narrow(axis, start, min(step, length-start)), block_bytes) + return + raise CacheBypass("unrepresentable tensor block") + + +def tensor_digest(tensor, interrupt=None): + h = hashlib.sha256() + for block in tensor_blocks(tensor): + if interrupt is not None: + interrupt() + h.update(block) + return h.hexdigest() + + +def canonical(value: Any, interrupt=None, depth=0): + """Reject opaque objects instead of using an unreliable repr or pointer key.""" + if depth > 16: + raise CacheBypass("metadata nesting limit") + if value is None or type(value) in (str, bool, int): + return value + if type(value) is float: + if not math.isfinite(value): + raise CacheBypass("non-finite metadata") + return value + if isinstance(value, (torch.dtype, torch.device)): + return str(value) + if isinstance(value, torch.Tensor): + return ["tensor", str(value.dtype), list(value.shape), tensor_digest(value, interrupt)] + if isinstance(value, Mapping): + if len(value) > 256 or any(not isinstance(k, str) for k in value): + raise CacheBypass("unsupported metadata mapping") + return {k: canonical(value[k], interrupt, depth+1) for k in sorted(value)} + if isinstance(value, (tuple, list)) and len(value) <= 4096: + return [canonical(x, interrupt, depth+1) for x in value] + raise CacheBypass(f"opaque metadata: {type(value).__name__}") + + +def selected_latent(samples, stream): + latent = samples["samples"] + if getattr(latent, "is_nested", False): + parts = latent.unbind() + latent = parts[0 if stream == "video" else -1] + if not isinstance(latent, torch.Tensor): + raise CacheBypass("unknown latent payload") + return latent + + +def _callable_token(function): + function = getattr(function, "__func__", function) + return TOKENS.token(function) + + +def _empty_option_tree(value): + if value is None: + return True + if isinstance(value, Mapping): + return all(_empty_option_tree(v) for v in value.values()) + if isinstance(value, (tuple, list)): + return all(_empty_option_tree(v) for v in value) + return False + + +def vae_signature(vae, stream, delegate): + """Supported native H3 inference VAE only; unknown wrappers still decode normally. + + This is not a model-file hash. Unsupported in-place .data changes require + reset_token or a reloaded VAE. No weights are copied or read onto the CPU. + """ + model = getattr(vae, "first_stage_model", None) + patcher = getattr(vae, "patcher", None) + if type(vae).__module__ != "comfy.sd" or not type(model).__module__.startswith("comfy.ldm.minimax."): + raise CacheBypass("unrecognized VAE identity; native decode only") + if patcher is None or not hasattr(patcher, "patches_uuid"): + raise CacheBypass("VAE patch version unavailable") + if bool(getattr(model, "training", True)): + raise CacheBypass("training-mode VAE") + if "decode" in vars(vae) or "decode" in vars(model): + raise CacheBypass("instance-patched decoder") + if getattr(patcher, "object_patches", None): + raise CacheBypass("VAE object patches require cache bypass") + + if not _empty_option_tree(getattr(patcher, "model_options", None)): + raise CacheBypass("non-default VAE patcher options") + if any(getattr(patcher, field, None) for field in ("forced_hooks", "hook_patches")): + raise CacheBypass("VAE hook state requires cache bypass") + if any(getattr(part, "_forward_hooks", None) or getattr(part, "_forward_pre_hooks", None) + for part in model.modules()): + raise CacheBypass("VAE forward hooks require cache bypass") + + config = {} + for name in ("vae_dtype", "device", "output_device", "audio_sample_rate", "audio_sample_rate_output", + "latent_channels", "disable_offload", "output_channels"): + if hasattr(vae, name): + config[name] = canonical(getattr(vae, name)) + # Native H3 tiling/temporal configuration is scalar, not model weights. + model_config = {k: v for k, v in vars(model).items() + if not k.startswith("_") and type(v) in (str, bool, int, float, type(None))} + # Model residency can replace buffer objects without changing their values. + # Logical names/versions + Core patch UUID are used, not buffer pointers. + # Reloading weights through raw .data (outside Core) requires reset_token. + versions = [] + for name, value in itertools.chain(model.named_parameters(), model.named_buffers()): + try: + version = value._version + except RuntimeError: # inference tensors intentionally have no counter + version = None + versions.append((name, version, str(value.dtype), tuple(value.shape))) + identity = ( + TOKENS.token(vae), TOKENS.token(model), TOKENS.token(patcher), + str(patcher.patches_uuid), _callable_token(vae.decode), _callable_token(model.decode), + _callable_token(delegate), + _callable_token(vae.process_output) if callable(getattr(vae, "process_output", None)) else None, + config, model_config, versions, + str(torch.get_default_dtype()), torch.get_float32_matmul_precision(), + bool(torch.backends.cuda.matmul.allow_tf32), bool(torch.backends.cudnn.allow_tf32), + bool(torch.backends.cudnn.benchmark), bool(torch.backends.cudnn.deterministic), + bool(torch.are_deterministic_algorithms_enabled()), + bool(torch.is_grad_enabled()), bool(torch.is_inference_mode_enabled()), + bool(torch.is_autocast_enabled("cpu")), str(torch.get_autocast_dtype("cpu")), + bool(torch.is_autocast_enabled("cuda")), str(torch.get_autocast_dtype("cuda")), + ) + payload = json.dumps(identity, sort_keys=True, separators=(",", ":"), allow_nan=False) + return hashlib.sha256(payload.encode()).hexdigest(), (weakref.ref(vae), weakref.ref(model), weakref.ref(patcher)) + + +def make_key(samples, stream, signature, reset_token=0, interrupt=None): + latent = selected_latent(samples, stream) + # Only the chosen AV stream is hashed. Other decode-relevant metadata, + # notably samples['sample_rate'], is included, conservatively, as well. + metadata = {k: v for k, v in samples.items() if k != "samples"} + payload = [CONTRACT, stream, signature, int(reset_token), str(latent.dtype), list(latent.shape), + tensor_digest(latent, interrupt), canonical(metadata, interrupt)] + return hashlib.sha256(json.dumps(payload, sort_keys=True, separators=(",", ":")).encode()).hexdigest() diff --git a/decode_cache/nodes.py b/decode_cache/nodes.py new file mode 100644 index 0000000..ce41683 --- /dev/null +++ b/decode_cache/nodes.py @@ -0,0 +1,118 @@ +"""One independent helper; existing Continuum node IDs and schemas are untouched.""" +from __future__ import annotations + +import json +import logging +import threading + +from .service import DecodeService, reraise_control +from .store import MiB, GiB + + +class H3DecodeCacheHelper: + CATEGORY = "MiniMax H3/Continuum/Helpers" + FUNCTION = "decode" + RETURN_TYPES = ("IMAGE", "AUDIO", "STRING") + RETURN_NAMES = ("images", "audio", "report") + INPUT_IS_LIST = True + OUTPUT_IS_LIST = (True, True, False) + DESCRIPTION = ( + "Optional V3.8X2 helper: reuse complete physical VAE decode entries. " + "Auto stores video in a private process-local disk cache, audio in bounded RAM. " + "Connect the unchanged Assembly Plan directly to Finalize. No model unload or sampling changes." + ) + SEARCH_ALIASES = ["Decode Cache Helper", "H3 cached decode", "video audio cache"] + + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "cache_mode": (["Auto", "RAM", "Off"], {"default": "Auto", "tooltip": + "Auto: disk-backed video + small RAM audio. RAM: bounded RAM only. Off: native decoding, clear helper cache."}), + "ram_budget_mb": ("INT", {"default": 256, "min": 0, "max": 8192, "step": 128, "advanced": True, + "tooltip": "Additional private RAM retention limit (MiB), not total process memory. Low headroom disables RAM retention."}), + "disk_budget_gb": ("INT", {"default": 8, "min": 0, "max": 128, "step": 1, "advanced": True, + "tooltip": "Process-local cache quota in GiB. Set 0 for no cache file writes. Auto does not reuse files after restart."}), + "reset_token": ("INT", {"default": 0, "min": 0, "max": 2147483647, "advanced": True, + "tooltip": "Increase to discard this helper's cache, especially after unsupported in-place VAE weight edits."}), + }, + "optional": { + "video_samples": ("LATENT", {"tooltip": "Continuum video_latents LIST; each entry is decoded whole, before trim/seam."}), + "video_vae": ("VAE", {"tooltip": "Same native Video VAE as the existing VAE Decode node."}), + "audio_samples": ("LATENT", {"tooltip": "Continuum audio_latents LIST; optional, decoded independently."}), + "audio_vae": ("VAE", {"tooltip": "Same native Audio VAE as VAE Decode Audio. Required only with audio_samples."}), + }, + } + + @classmethod + def IS_CHANGED(cls, **kwargs): + # Review must still execute; only this helper's content cache skips decode. + return float("nan") + + def __init__(self): + self.service = DecodeService() + self._lock = threading.RLock() + + @staticmethod + def _scalar(value, name): + if isinstance(value, list): + if len(value) != 1: + raise ValueError(f"{name} expects one value, not a list batch") + return value[0] + return value + + @staticmethod + def _list(value): + return value if isinstance(value, list) else ([] if value is None else [value]) + + def _stream(self, samples, vaes, stream, reset_token): + entries = self._list(samples) + vaes = self._list(vaes) + if not entries: + return [], [] + outputs, events = [], [] + for index, entry in enumerate(entries): + if entry is None: + # Same missing-output position; do not duplicate the previous entry. + outputs.append(None) + events.append({"stream": stream, "entry": index, "status": "unconnected"}) + continue + vae = vaes[min(index, len(vaes)-1)] if vaes else None + result, event = self.service.decode(entry, vae, stream, reset_token) + outputs.append(result) + events.append({"entry": index, **event}) + return outputs, events + + def decode(self, cache_mode, ram_budget_mb, disk_budget_gb, reset_token, + video_samples=None, video_vae=None, audio_samples=None, audio_vae=None): + with self.service.store.operation(), self._lock: + mode = self._scalar(cache_mode, "cache_mode") + ram = int(self._scalar(ram_budget_mb, "ram_budget_mb")) + disk = int(self._scalar(disk_budget_gb, "disk_budget_gb")) + reset = int(self._scalar(reset_token, "reset_token")) + if mode not in ("Auto", "RAM", "Off") or ram < 0 or disk < 0: + raise ValueError("invalid cache settings") + config_error = None + try: + self.service.store.configure(mode, ram*MiB, disk*GiB, reset) + except Exception as exc: + reraise_control(exc) + self.service.store.mode = "Off" + config_error = str(exc)[:160] + logging.getLogger("h3_decode_cache_helper").warning("Cache unavailable; native decode: %s", exc) + images, video_events = self._stream(video_samples, video_vae, "video", reset) + audio, audio_events = self._stream(audio_samples, audio_vae, "audio", reset) + try: + stats = self.service.store.stats() + except Exception as exc: + reraise_control(exc) + stats = {"diagnostic_error": str(exc)[:160]} + report = {"helper_version": "3.8.3", "mode": mode, + "video": video_events, "audio": audio_events, "cache": stats} + if config_error: + report["cache_disabled_reason"] = config_error + return images, audio, json.dumps(report, ensure_ascii=False, indent=2) + + +NODE_CLASS_MAPPINGS = {"H3DecodeCacheHelper": H3DecodeCacheHelper} +NODE_DISPLAY_NAME_MAPPINGS = {"H3DecodeCacheHelper": "Decode Cache Helper"} diff --git a/decode_cache/service.py b/decode_cache/service.py new file mode 100644 index 0000000..4ae715d --- /dev/null +++ b/decode_cache/service.py @@ -0,0 +1,115 @@ +"""Native decoder delegation. Cache errors fail open; decoder/cancellation errors do not.""" +from __future__ import annotations + +import importlib +import json +import logging +import time + + +from .identity import make_key, vae_signature +from .store import DecodeStore + +LOG = logging.getLogger("h3_decode_cache_helper") + + +def interrupt_check(): + try: + mm = importlib.import_module("comfy.model_management") + except ImportError: + return + mm.throw_exception_if_processing_interrupted() + + +def reraise_control(exc): + # ComfyUI's interruption is an Exception, unlike KeyboardInterrupt. + # Never convert an interrupted Queue into another Core Decode attempt. + if any(cls.__name__ == "InterruptProcessingException" for cls in type(exc).__mro__): + raise exc + + +def native_delegate(stream): + if stream == "video": + core = importlib.import_module("nodes").VAEDecode + method = core.decode + + def execute(vae, samples): + return core().decode(vae=vae, samples=samples)[0] + return method, execute + helper = importlib.import_module("comfy_extras.nodes_audio").vae_decode_audio + return helper, lambda vae, samples: helper(vae, samples) + + +class DecodeService: + def __init__(self, store=None, delegate_resolver=native_delegate, + signature_fn=vae_signature, interrupt=interrupt_check): + self.store = store or DecodeStore(interrupt=interrupt) + self.delegate_resolver = delegate_resolver + self.signature_fn = signature_fn + self.interrupt = interrupt + + def decode(self, samples, vae, stream, reset_token=0): + self.interrupt() + delegate_identity, native = self.delegate_resolver(stream) + event = {"stream": stream, "status": "miss", "hash_s": 0.0, + "read_s": 0.0, "decode_s": 0.0, "write_s": 0.0} + signature = key = guards = None + epoch = self.store.epoch + if self.store.mode != "Off": + start = time.perf_counter() + try: + signature, guards = self.signature_fn(vae, stream, delegate_identity) + key = make_key(samples, stream, signature, reset_token, self.interrupt) + except Exception as exc: + reraise_control(exc) + event["status"] = "bypass" + event["reason"] = str(exc)[:160] + event["hash_s"] = time.perf_counter() - start + else: + event["status"] = "off" + if key is not None: + start = time.perf_counter() + try: + cached = self.store.get(key) + if cached is not None: + tensor, extra = cached + event.update(status="hit", read_s=time.perf_counter()-start) + return (tensor if stream == "video" else {"waveform": tensor, **extra}), event + except Exception as exc: + reraise_control(exc) + event["cache_read_error"] = str(exc)[:160] + LOG.warning("Decode cache read failed; using native decoder: %s", exc) + event["read_s"] = time.perf_counter() - start + + # Outside cache exception handlers. No blind retry of OOM, invalid + # inputs, VAE errors, or interrupted Sampling/Decode. + self.interrupt() + start = time.perf_counter() + decoded = native(vae, samples) + event["decode_s"] = time.perf_counter() - start + self.interrupt() + if key is not None: + start = time.perf_counter() + try: + after, _ = self.signature_fn(vae, stream, delegate_identity) + if after != signature: + event["store_skipped"] = "VAE identity changed during native decode" + else: + if stream == "video": + tensor, extra = decoded, {} + else: + tensor = decoded["waveform"] + extra = {k: v for k, v in decoded.items() if k != "waveform"} + # Only JSON-compatible metadata is retained; store defensively copies it. + json.dumps(extra, allow_nan=False) + replacement = self.store.put(key, tensor, stream, extra, guards, epoch) + event["stored"] = self.store.contains(key) + if replacement is not None: + tensor, extra = replacement + decoded = tensor if stream == "video" else {"waveform": tensor, **extra} + except Exception as exc: + reraise_control(exc) + event["cache_write_error"] = str(exc)[:160] + LOG.warning("Decode cache write failed; returning native result: %s", exc) + event["write_s"] = time.perf_counter() - start + return decoded, event diff --git a/decode_cache/store.py b/decode_cache/store.py new file mode 100644 index 0000000..f06c785 --- /dev/null +++ b/decode_cache/store.py @@ -0,0 +1,427 @@ +"""Private, disposable decoded tensors. No Run Storage or global GPU lifecycle access.""" +from __future__ import annotations + +from copy import deepcopy +import hashlib +import os +from pathlib import Path +import shutil +import tempfile +import threading +import uuid +import weakref +from contextlib import nullcontext +from collections import OrderedDict +from dataclasses import dataclass, field + +import numpy as np +import torch + +from .identity import CacheBypass, tensor_blocks, tensor_digest + +if "." in __package__: + from .. import storage_settings, storage_fs + from ..storage_memory import system_memory_bytes +else: # The existing standalone cache test/probe entry points remain supported. + storage_settings = storage_fs = None + system_memory_bytes = None + +MiB = 1024**2 +GiB = 1024**3 + + +def available_ram(): + if system_memory_bytes is not None: + return system_memory_bytes()[0] or 0 + try: + import psutil # already provided by ComfyUI; optional for cache operation + return int(psutil.virtual_memory().available) + except (ImportError, OSError, AttributeError): + return 0 # unknown headroom must not trigger RAM retention + + +@dataclass +class Entry: + shape: tuple + dtype: str + digest: str + payload_bytes: int + guards: tuple + extra: dict + tensor: torch.Tensor | None = None + path: Path | None = None + file_bytes: int = 0 + leases: list = field(default_factory=list) + + def live(self): + self.leases[:] = [ref for ref in self.leases if ref() is not None] + return bool(self.leases) + + +def _cleanup_private_files(state): + """Best effort at object/process exit; never close arrays still owned downstream.""" + with state.get("lock", nullcontext()): + _cleanup_private_files_locked(state) + + +def _cleanup_private_files_locked(state): + for path, entry in list(state["entries"].items()): + if not entry.live(): + try: + if path.parent in state.get("identities", {}): + storage_fs.check_identity(path.parent, state["identities"][path.parent]) + path.unlink(missing_ok=True) + state["entries"].pop(path, None) + except OSError: + pass + for root, owner in state.get("owners", {}).items(): + try: + storage_fs.check_identity(root, owner["identity"]) + for name in owner["files"]: + path = root / name + if path not in state["entries"] and storage_fs.regular_file(path): + try: + path.unlink() + except OSError: + pass + except OSError: + pass + for root, expected in state.get("identities", {}).items(): + storage_settings.remove_empty_cache_root(root, expected) + if state["root"] is not None and state["root"] not in state.get("identities", {}): + try: + state["root"].rmdir() + except OSError: + pass + + +class DecodeStore: + """One store per Helper instance. Auto uses disk for video and small RAM for audio. + + RAM hits return a clone. Disk hits return a new copy-on-write mmap, never + a writable shared backing file. Already-returned results outlive eviction. + """ + + def __init__(self, temp_parent=None, memory_probe=available_ram, disk_probe=shutil.disk_usage, + headroom_bytes=2*GiB, interrupt=None): + self._lock = threading.RLock() + self.entries = OrderedDict() + self.retired = [] + self.root = None + self.temp_parent = temp_parent if temp_parent is not None else (os.environ.get("H3_DECODE_CACHE_TEMP_DIR") or None) + self._explicit_parent = temp_parent is not None + self._default_parent = self.temp_parent + self._owners = {} + self._scratch_token = uuid.uuid4().hex + self.memory_probe = memory_probe + self.disk_probe = disk_probe + self.headroom_bytes = headroom_bytes + self.interrupt = interrupt or (lambda: None) + self.ram_limit = 512*MiB + self.disk_limit = 16*GiB + self.mode = "Auto" + self.reset_token = 0 + self.epoch = 0 + self.max_entries = 256 + # Finalizer state has only paths, disk-entry metadata and weak leases; + # no decoded RAM tensor, VAE, store, node, or GPU model is retained. + self._cleanup_state = {"root": None, "entries": {}, "identities": {}, + "lock": self._lock, "owners": self._owners} + weakref.finalize(self, _cleanup_private_files, self._cleanup_state) + + def operation(self): + manager = storage_settings.get_manager() if storage_settings else None + return manager.operation() if manager else nullcontext() + + def configure(self, mode, ram_limit, disk_limit, reset_token): + with self._lock: + settings = (mode, int(ram_limit), int(disk_limit), int(reset_token)) + if settings != (self.mode, self.ram_limit, self.disk_limit, self.reset_token): + self.clear() + self.mode, self.ram_limit, self.disk_limit, self.reset_token = settings + self._reap() + if self.memory_probe() < self.headroom_bytes: + for key, entry in list(self.entries.items()): + if entry.tensor is not None: + self._drop(key) + + @property + def ram_bytes(self): + return sum(e.payload_bytes for e in self.entries.values() if e.tensor is not None) + + @property + def disk_bytes(self): + return sum(e.file_bytes for e in self.entries.values() if e.path is not None) + sum(e.file_bytes for e in self.retired) + + def _delete(self, entry): + if entry.path is None: + return True + if entry.live(): + return False + try: + if entry.path.parent in self._owners: + storage_fs.check_identity(entry.path.parent, self._owners[entry.path.parent]["identity"]) + entry.path.unlink(missing_ok=True) + self._cleanup_state["entries"].pop(entry.path, None) + return True + except OSError: + return False + + def _drop(self, key): + entry = self.entries.pop(key) + if entry.path is not None and not self._delete(entry): + self.retired.append(entry) + + def _reap(self): + self.retired[:] = [e for e in self.retired if not self._delete(e)] + for key, entry in list(self.entries.items()): + if not all(ref() is not None for ref in entry.guards): + self._drop(key) + + def clear(self): + with self._lock: + self.epoch += 1 + for key in list(self.entries): + self._drop(key) + self._reap() + + def close(self): + self.clear() + for root, owner in self._owners.items(): + storage_settings.remove_empty_cache_root(root, owner["identity"]) + if self.root is not None: + try: + self.root.rmdir() # only an empty helper-owned directory + except OSError: + pass + + def _directory(self): + manager = storage_settings.get_manager() if storage_settings else None + if manager is not None: + manager.register_store(self) + default = self._default_parent or tempfile.gettempdir() + parent = Path(default).resolve() if self._explicit_parent else manager.resolve("decode", default) + if self.root is not None and self.root.parent != parent: + self.root = None + self.temp_parent = parent + if self.root is None: + self.root = Path(tempfile.mkdtemp(prefix=f"h3-decode-cache-{os.getpid()}-", dir=self.temp_parent)) + self._cleanup_state["root"] = self.root + if manager is not None: + owner = storage_settings.create_cache_owner(self.root) + self._owners[self.root] = owner + self._cleanup_state["identities"][self.root] = owner["identity"] + manager.record_root("decode", self.root.parent) + if self.root in self._owners: + storage_fs.check_identity(self.root, self._owners[self.root]["identity"]) + return self.root + + def _ram_room(self, size): + while self.ram_bytes + size > self.ram_limit: + victim = next((k for k, e in self.entries.items() if e.tensor is not None), None) + if victim is None: + return False + self._drop(victim) + # Budget is additional retention only, not a claim about whole-process RSS. + return self.memory_probe() >= self.headroom_bytes + 2*size + + def _disk_room(self, size): + self._reap() + while self.disk_bytes + size > self.disk_limit: + victim = next((k for k, e in self.entries.items() if e.path is not None and not e.live()), None) + if victim is None: + return False + self._drop(victim) + root = self._directory() + return self.disk_probe(root).free >= size + 256*MiB + + def _read_disk(self, entry): + path = entry.path + if path is not None and path.parent in self._owners: + storage_fs.check_identity(path.parent, self._owners[path.parent]["identity"]) + if path is None or path.is_symlink() or path.stat().st_size != entry.file_bytes: + raise CacheBypass("disk cache file missing/changed") + # Verify the exact serialized payload before exposing a mapping. + # Header/size are also checked; pickle is never allowed. + array = np.load(path, mmap_mode="c", allow_pickle=False, max_header_size=4096) + if tuple(array.shape) != entry.shape or array.dtype.str != entry.dtype or not array.flags.c_contiguous: + raise CacheBypass("disk cache header mismatch") + if array.nbytes != entry.payload_bytes: + raise CacheBypass("disk cache size mismatch") + raw = memoryview(array).cast("B") + h = hashlib.sha256() + for start in range(0, len(raw), 8*MiB): + self.interrupt() + h.update(raw[start:start+8*MiB]) + if h.hexdigest() != entry.digest: + raise CacheBypass("disk cache checksum mismatch") + entry.leases.append(weakref.ref(array)) + # mode='c' makes writable, private pages. mode='r' is NOT used. + return torch.from_numpy(array) + + def get(self, key): + with self._lock: + self._reap() + entry = self.entries.get(key) + if entry is None: + return None + if not all(ref() is not None for ref in entry.guards): + self._drop(key) + return None + self.entries.move_to_end(key) + if entry.tensor is not None: + if self.memory_probe() < self.headroom_bytes + entry.payload_bytes: + self._drop(key) + return None + if (tuple(entry.tensor.shape) != entry.shape + or entry.tensor.numpy().dtype.str != entry.dtype + or tensor_digest(entry.tensor, self.interrupt) != entry.digest): + self._drop(key) + raise CacheBypass("RAM cache checksum/header mismatch") + # Independent storage protects the cache from in-place seam/color edits. + return entry.tensor.clone(), deepcopy(entry.extra) + try: + tensor = self._read_disk(entry) + except Exception: + self._drop(key) + raise + return tensor, deepcopy(entry.extra) + + def put(self, key, tensor, stream, extra, guards, expected_epoch): + with self._lock: + if expected_epoch != self.epoch or self.mode == "Off": + return None + if (not isinstance(tensor, torch.Tensor) or tensor.device.type != "cpu" + or tensor.layout != torch.strided or tensor.is_nested or tensor.numel() == 0): + return None + if tensor.dtype not in (torch.float16, torch.float32, torch.float64): + return None # never silently downcast decoded output + size = tensor.numel() * tensor.element_size() + if not all(ref() is not None for ref in guards): + return None + if key in self.entries: + self._drop(key) + while len(self.entries) >= self.max_entries: + self._drop(next(iter(self.entries))) + use_ram = self.mode == "RAM" or (stream == "audio" and size <= 16*MiB) + if use_ram and self._ram_room(size): + stored = tensor.detach().clone(memory_format=torch.contiguous_format) + self.entries[key] = Entry(tuple(stored.shape), stored.numpy().dtype.str, + tensor_digest(stored, self.interrupt), size, guards, deepcopy(extra), tensor=stored) + return None # original MISS output is independently owned + if self.mode != "Auto" or self.disk_limit == 0 or size+4096 > self.disk_limit: + return None + if not self._disk_room(size+4096): + return None + root = self._directory() + filename = uuid.uuid4().hex + if root in self._owners: + storage_settings.register_cache_file(root, self._owners[root], filename) + temp = root / (filename + ".tmp") + final = root / (filename + ".npy") + dtype = {torch.float16: np.dtype("float16"), torch.float32: np.dtype("float32"), + torch.float64: np.dtype("float64")}[tensor.dtype] + h = hashlib.sha256() + try: + with open(temp, "xb") as f: + np.lib.format.write_array_header_1_0(f, { + "descr": dtype.str, "fortran_order": False, "shape": tuple(tensor.shape)}) + for block in tensor_blocks(tensor): + self.interrupt() + h.update(block) + if f.write(block) != len(block): + raise OSError("short cache write") + f.flush() + os.fsync(f.fileno()) + os.replace(temp, final) + entry = Entry(tuple(tensor.shape), dtype.str, h.hexdigest(), size, guards, + deepcopy(extra), path=final, file_bytes=final.stat().st_size) + self.entries[key] = entry + self._cleanup_state["entries"][final] = entry + # Returning a COW mapping releases the large anonymous MISS output + # after this call rather than keeping an extra private RAM copy. + return self.get(key) + except BaseException: + if key in self.entries: + self._drop(key) + for path in (temp, final): + try: + path.unlink(missing_ok=True) + except OSError: + pass + raise + + def contains(self, key): + with self._lock: + return key in self.entries + + def inspect_scratch(self, *, cleanup=False, parent): + """Pure status unless explicitly cleaning; preserve aliases and RAM entries.""" + with self._lock: + result = {} + removed = {"removed_files": 0, "removed_bytes": 0, "errors": []} + entries = [(key, entry) for key, entry in self.entries.items()] + entries += [(None, entry) for entry in self.retired] + for key, entry in entries: + path = entry.path + if path is None or path.parent.parent != Path(parent): + continue + try: + if path.parent in self._owners: + storage_fs.check_identity(path.parent, self._owners[path.parent]["identity"]) + if not path.is_file() or path.is_symlink(): + continue + size = path.stat().st_size + live = any(ref() is not None for ref in entry.leases) + if cleanup and not live and self._delete(entry): + if key is not None: + self.entries.pop(key, None) + else: + self.retired = [item for item in self.retired if item is not entry] + removed["removed_files"] += 1 + removed["removed_bytes"] += size + else: + result[str(path)] = "in_use" if live else "deletable" + if cleanup and not live: + removed["errors"].append(f"Could not delete {path.name}; file retained") + except OSError as exc: + result[str(path)] = "unknown" + if cleanup: + removed["errors"].append(str(exc)) + # An interrupted unpublished write can leave a registered .tmp. + # The store lock proves no writer is active here; the entry ledger + # continues to protect any published mapping, including retired ones. + for root, owner in self._owners.items(): + if root.parent != Path(parent): + continue + try: + storage_fs.check_identity(root, owner["identity"]) + for name in owner["files"]: + path = root / name + if path in self._cleanup_state["entries"] or not storage_fs.regular_file(path): + continue + size = path.stat().st_size + if cleanup: + try: + path.unlink() + removed["removed_files"] += 1 + removed["removed_bytes"] += size + except OSError as exc: + result[str(path)] = "deletable" + removed["errors"].append(str(exc)) + else: + result[str(path)] = "deletable" + except OSError: + pass + if cleanup: + for root, owner in self._owners.items(): + if root.parent == Path(parent) and root != self.root: + storage_settings.remove_empty_cache_root(root, owner["identity"]) + result[f"__removed_{self._scratch_token}"] = removed + return result + + def stats(self): + with self._lock: + self._reap() + return {"entries": len(self.entries), "ram_bytes": self.ram_bytes, + "disk_bytes": self.disk_bytes, "pending_delete_files": len(self.retired)} diff --git a/docs/AB_TEST_LOG_JA.md b/docs/AB_TEST_LOG_JA.md new file mode 100644 index 0000000..bf72bba --- /dev/null +++ b/docs/AB_TEST_LOG_JA.md @@ -0,0 +1,291 @@ +# H3 Continuum A/B実機テスト記録 + +更新日: 2026-08-13 +状態: 継続中。現時点の結果であり、最終推奨ではない。 + +## 1. 目的 + +H3 Continuumの3×5秒生成について、次を分離して評価する。 + +- 解像度とチャンク累積による顔・肌・Identity drift +- Spectrumの速度向上とSeed依存のForecast誤差 +- WarmupとContinuum Actual Prefixの組み合わせ +- Seam Correction Off/Autoの映像・音声境界への効果 +- 無音に近い条件と明確な歌唱・発話条件の違い + +MP4本体は現時点ではGitへ追加しない。各番号はローカル出力ファイル名の末尾番号に対応する。 + +## 2. 主なテスト環境 + +```text +GPU RTX 5060 Ti 16GB +System RAM 64GB +Output 640×960 / 24fps / 15秒 / 360 frames +Image scale area / 0.60MP / resolution_steps 32 +Chunks 3 +Chunk seconds 5.0 +Continuity Balanced — 22 frames +SageAttention sageattn_qk_int8_pv_fp8_cuda++ +Sol-Attn なし +Audio continuity true +Exact total duration true +Diagnostics Basic +``` + +Spectrum ON時の主要設定: + +```text +blend_weight 0.5 +degree 1 +ridge_lambda 0.1 +window_size 2.0 +flex_window 0.75 +tail_actual_steps 1 +max_history 8 +history_storage system_ram +bootstrap_first_forecast true +anchor_residual_feedback false +selective_rollback_correction false +offline_smoothing_replay true +audio_blend_weight 0.0 +``` + +## 3. 暫定実用設定 + +品質最優先ではなく、約31%の時間短縮を取る実用バランス設定として扱う。 + +```text +Spectrum On +Warmup 2 +Actual Prefix 2 +Continuity Balanced 22 +Seam Off +``` + +注意: Seed 1235でSpectrum ON時の明確な画質崩れを確認しているため、全Seedで安定する設定ではない。 + +## 4. 解像度・チャンク比較 + +| 動画 | 解像度 | 尺 | 所見 | +|---|---:|---:|---| +| 00032 | 448×672、約0.3MP | 15秒 | 低解像度では肌・顔の微細なdriftが目立ちにくい | +| 00033 | 576×896、約0.5MP | 15秒 | 比較的安定 | +| 00034 | 576×896、約0.5MP | 5秒 | 単チャンクは安定 | +| 00035 | 640×960、約0.6MP | 5秒 | 0.6MP単独では大きな破綻なし | +| 00036 | 640×960、約0.6MP | 15秒 | 後続チャンクほど肌の変化が見えやすい | + +暫定解釈: 0.6MPそのものより、Continuationで小さな誤差が次チャンクへ累積し、高解像度で可視化される可能性が高い。 + +## 5. Warmup比較 + +| 動画 | Warmup | 所見 | +|---|---:|---| +| 00038 | 1 | プッシュインあり。境界フリックは比較的目立たない | +| 00039 | 3 | 6秒台以降に顔の暗転・粗れが発生し、重大な破綻 | +| 00040 | 3 | 重大破綻は再現せず、軽い肌drift。結果の振れが大きい | +| 00041 | 2 | Seed 1234、Spectrum ON、Seam Off。全体として良好 | + +暫定解釈: + +- Warmup 3は結果が不安定なため不採用。 +- Warmup 2はWarmup 3より安全だが、Spectrumの全Seed安定性は保証しない。 +- Continuation chunkではActual Prefix 2が先頭2stepをActualへ要求する。 + +## 6. 厳密または準厳密なA/B比較 + +境界比は、境界の隣接フレームMAEを周辺の通常フレーム間MAE中央値で割った参考値。1未満なら境界変化が周辺動作より小さい。 + +### 6.1 Spectrum ON/OFF、Seed 1234 + +| 動画 | Spectrum | Warmup | Seam | 所見 | +|---|---|---:|---|---| +| 00041 | ON | 2 | Off | 良好。音声はほぼ無音、RMS約-63.7dBFS | +| 00045 | OFF | 無効 | Off | 良好。00041と視覚的にかなり近い | + +Seed 1234ではSpectrum ONでもNativeに近い結果を維持した。 + +### 6.2 Spectrum ON/OFF、Seed 1235 + +| 動画 | Spectrum | Warmup | Seam | 5秒比 | 10秒比 | 所見 | +|---|---|---:|---|---:|---:|---| +| 00042 | ON | 2 | Off | 1.392 | 1.810 | Clip 1途中で服色変化。Clip 3で顔に強い斑点と肌崩れ | +| 00047 | OFF | 無効 | Off | 0.909 | 0.904 | 服色、肌、Identityを維持 | + +確定事項: + +- Seed 1235自体はNative生成で正常。 +- 00042の崩れはSpectrumとSeed 1235の組み合わせで発生した可能性が高い。 +- Seam CorrectionではClip内部のSpectrum Forecast誤差を修正できない。 + +### 6.3 Seam Auto/Off、Seed 1237 + +| 動画 | Spectrum | Warmup | Seam | 5秒比 | 10秒比 | 所見 | +|---|---|---:|---|---:|---:|---| +| 00044 | ON | 2 | Auto | 0.570 | 0.248 | 良好。Autoによる不自然な色変化なし | +| 00046 | ON | 2 | Off | 0.564 | 0.333 | 良好。Auto版と視覚差は非常に小さい | + +確定事項: + +- 無音に近い映像ではAutoの視覚的効果は小さい。 +- Autoは10秒境界をわずかに平滑化したが、Offでも十分安定。 +- 現時点ではOffを映像品質の基準とする。 + +## 7. Seed耐性 + +| Seed | Spectrum | Warmup | Seam | 動画 | 判定 | +|---:|---|---:|---|---|---| +| 1234 | ON | 2 | Off | 00041 | 良好 | +| 1234 | OFF | 無効 | Off | 00045 | 良好 | +| 1235 | ON | 2 | Off | 00042 | 失敗 | +| 1235 | OFF | 無効 | Off | 00047 | 良好 | +| 1236 | ON | 2 | Auto | 00043 | 許容。カメラ動作が強い | +| 1237 | ON | 2 | Auto | 00044 | 良好 | +| 1237 | ON | 2 | Off | 00046 | 良好 | + +Spectrum ONの品質はSeed依存で変化する。結果確認なしの長尺・大量生成には追加の誤差検出またはフォールバックが必要。 + +## 8. 速度比較 + +| 条件 | 動画 | 生成時間 | +|---|---|---:| +| Spectrum ON | 00046 | 882.52秒、約14分43秒 | +| Spectrum OFF | 00047 | 1274.40秒、約21分14秒 | + +```text +短縮時間 約391.88秒、約6分32秒 +短縮率 約30.8% +処理速度 約1.44倍 +``` + +別の比較でも約31.6%短縮しており、Spectrumの速度効果は再現している。ただし00046と00047はSeedが異なるため、この値は実用上の参考値。 + +## 9. 音声の現状 + +00041〜00047にはAAC音声ストリームがあるが、測定RMSは概ね-58〜-68dBFSで、聴感上ほぼ無音。したがって、歌唱、会話、伴奏を含むAudio ContinuityとSeam Autoの効果は未確定。 + +現時点の運用仮説: + +```text +映像中心・無音に近い Seam Off +会話・歌唱・演奏あり Seam Autoの可能性。A/B未完了 +``` + +## 10. 保留中・次のテスト + +1. Seed 1235 / Spectrum ON / Warmup 1 / Seam Off +2. 固定カメラ・明確な持続歌唱で、同一SeedのSeam Off/Auto +3. 歌唱と小振幅・低速プッシュインを含む実用Promptで、同一SeedのSeam Off/Auto +4. 音声境界の前後100〜200msについて、瞬間ピーク、短時間RMS、F0、波形相関、スペクトル差、DC offsetを比較 +5. 3チャンクで安定後、5×5秒または6×5秒でIdentity driftを確認 + +## 11. 検討中の対策 + +- Clip 1だけNative計算し、Continuation chunkからSpectrumを使用 +- Spectrumの予測誤差が大きいstepだけActualへ戻す +- selective rollback correctionとanchor residual feedbackの同一Seed A/B +- 元画像による定期的なIdentity再アンカー +- 異常Contextを検出し、該当チャンクだけSpectrumなしで再生成 +- Seam Autoの映像補正と音声補正を個別に選択可能にする + +## 12. 将来のGitHub公開方法 + +- このMarkdownと機械可読なJSON/CSVを通常Gitへ追加する。 +- コンタクトシートは`docs/assets/ab-tests/`等へ追加する。 +- 大容量MP4は通常Gitへ直接追加せず、GitHub Release assetsまたはGit LFSを検討する。 +- 公開前にPrompt、Workflow、モデル、ComfyUI revision、Python/Torch/CUDA、GPU、生成時間を各サンプルへ紐付ける。 + +## 00048 / 00049: 0.5 MP Seam Off vs Auto + +- 共通条件: 576x896 (約0.5 MP), 24 fps, 3 x 5秒, Spectrum On, Warmup 1, Actual Prefix 2, Balanced 22, base seed 1235。 +- 00048: Seam Off。5秒境界比 0.3583、10秒境界比 0.8562。 +- 00049: Seam Auto。5秒境界比 0.3709、10秒境界比 0.4271。 +- 視覚評価: 両方とも人物・衣装・肌は安定し、見た目の差は小さい。Autoは10秒境界を数値上改善した。 +- 音声: 両方ともほぼ無音のため、音声シーム効果は未評価。 +- 注意: 0.5 MP条件のため、0.6 MPで発生したSeed 1235の崩れをWarmup 1が解消したとは断定しない。 + +### 目視評価の補足 + +- ユーザー目視では00049(Seam Auto)の方がフリックが若干目立つ。 +- 00048(Seam Off)の方が境界変化を比較的抑えられている。 +- 境界指標の改善と主観的なフリック低減は一致しなかったため、本比較では目視評価を優先しSeam Offを採用候補とする。 + + +## 00050: 音声あり一貫性テスト + +- 条件: 576x896(約0.5 MP)、24 fps、3 x 5秒、Spectrum On、Warmup 1、Balanced 22、base seed 1235、Seam Off。 +- 音声レベル: mean -19.9 dB、peak -6.4 dB。前回のほぼ無音サンプルより音声境界の評価に適する。 +- 目視: 3チャンクを通して人物、髪、衣装、背景、露出は安定。肌・顔の累積崩れも目立たない。 +- 5秒境界: 境界付近で瞬きが始まるためフレーム差はあるが、露出フリックより自然動作として見える。 +- 10秒境界: 口形、顔、背景の連続性が高く、明確な視覚ジャンプは小さい。 +- 未確認: 音声境界のクリック、F0、短時間RMS、スペクトル差の数値解析。 +- 次の比較: 同一Prompt・Seed・解像度・Warmupのまま、SeamだけAutoへ変更した1本と比較する。 + +### 00050 音声の目視・聴感補足 + +- ユーザー聴感で10秒(Chunk 2→3境界)に短いプチノイズを確認。 +- 同境界の映像連続性は良好であり、映像シームと音声シームの品質が一致しない例となった。 +- 暫定対策候補: Video Seam Off、Audio Seam Autoを独立設定にする。 + + +### 00050 実行レポート補足 + +- Prompt plan: Fixed、3 chunks、5.000秒。Seam correction Off。 +- Chunk seeds: 12362262848839836333 / 1129406137053299874 / 8524677167168820503。 +- Continuity motion: Chunk 2 = 0.378492、Chunk 3 = 0.388277。Continuation chunksはActual Prefix 2を要求。 +- Retained frames: 124 + 119 + 119 = 362。Exact-durationで360 framesへ末尾調整。 +- Audio: 480000 samples(32 kHz、15.000秒)。末尾補正 -2667 samplesは2 frames相当で、10秒境界のプチノイズ原因とは分離して扱う。 +- 判定: sample-boundary alignmentは尺を整合させているが、波形位相・振幅の連続性までは保証しない。Audio Seam Off基準として保存する。 + +### 00050 境界評価の訂正 + +- 5秒(Chunk 1→2): 映像フリックが確認できる。 +- 10秒(Chunk 2→3): 映像フリックは目立たないが、音声に短いプチノイズがある。 +- 映像と音声の問題位置は一致していないため、Video SeamとAudio Seamは独立評価・独立処理が必要。 + +### 00050 最終訂正(この評価を優先) + +- 5秒(Chunk 1→2): 映像フリックあり。 +- 10秒(Chunk 2→3): 映像フリックは目立たない。 +- 音声: 明確なプチノイズは確認できない。先行する「10秒に音声ノイズあり」という記録は取り消し、未検出として扱う。 +- 00050はAudio Seam不具合の根拠にはせず、音声は専用テストで改めて検証する。 + +## V3.2.1 Reference + Continuation 実機成績 + +共通条件: + +```text +Version H3 Continuum V3.2.1 +Output 1056x608 / 24fps / 15秒 / 360 frames +Image scale 16:9 / 0.6MP / multiple 32 +Chunks 3 x 5秒 +Continuity Balanced 22 +Reference images 2 +Reference size Match Output +Audio Seam Off +Run Storage Save + Auto Resume +``` + +| 動画 | MODEL | Prompt | 生成時間 | Run Storage | 判定 | +|---|---|---|---:|---|---| +| 00006 | FL2VA | 縦構図指定が残る旧Prompt | 749.35秒 | 0 reused / 3 generated | MODEL切替前baseline | +| 00007 | Ref2VA | 縦構図指定が残る旧Prompt | 751.26秒 | 0 reused / 3 generated | Ref2VA切替を新Revisionとして検出 | +| 00008 | Ref2VA | 16:9横長へ修正したPrompt | 751.64秒 | 0 reused / 3 generated | 技術PASS、実用可能な結果 | + +00008の実行結果: + +```text +Conditioning mode Reference + Continuation +Prompt plan List / 3 unique prompts +Chunk 2 Actual Prefix 2 +Chunk 3 Actual Prefix 2 +Natural frames 362 +Final frames 360 +Revision 2167719e99b1e782 +``` + +確認事項: + +- 2枚のReferenceを全3チャンクで継続使用できた。 +- FL2VAからRef2VAへの変更、およびPrompt変更を別Revisionとして扱い、旧Chunkを誤再利用しなかった。 +- 00006~00008の生成時間は約749~752秒で、モデル切替やPrompt修正による異常な時間増加はない。 +- 00008はユーザー目視を含め、V3.2.1 Reference + Continuationの実用可能な実機成績として保存する。 diff --git a/docs/ARCHITECTURE.md b/docs/ARCHITECTURE.md new file mode 100644 index 0000000..bf973ea --- /dev/null +++ b/docs/ARCHITECTURE.md @@ -0,0 +1,177 @@ +# Architecture — H3 Continuum Join 2.0 + +## 1. Design goals + +1. Preserve the V1 native H3 continuation behavior. +2. Integrate N chunks without duplicating V1 temporal/layout logic. +3. Keep SageAttention, Sol-Attn, and Spectrum external and independently updatable. +4. Avoid global ComfyUI class mutation. +5. Fail closed when native H3 layout contracts change. +6. Prevent H3/VAE/CLIP model switching between continuation chunks. +7. Keep accepted chunk latents on CPU and bound transient decode memory. +8. Preserve exact long-chain A/V timing. + +## 2. Module layers + +```text +UI + nodes.py V1 compatibility nodes + v2/nodes.py integrated V2 nodes + +V2 orchestration + v2/sequence.py phases, session reuse, reroll, reports + v2/sampling.py ComfyUI-native sampler path + v2/h3_builder.py empty AV latent, prompt and keyframe conditioning + v2/decoder.py deferred decode, preallocated assembly + v2/prompts.py Fixed/List/Timeline parser + v2/session.py in-memory accepted chunk/session model + v2/session_io.py atomic session persistence + v2/motion.py conservative latent motion analysis + v2/seeds.py deterministic per-chunk seed derivation + v2/diagnostics.py lightweight seam measurements + +Shared continuation core + continuation.py native H3 video/video_audio reference construction + model_patch.py per-MODEL APPLY_MODEL wrapper + layout_adapter.py in-place MM-RoPE placement and payload normalization + temporal.py H3 17k+5 / 24fps / 40Hz grid + state.py V1 State and Plan schemas + media.py V1 trim/assembly helpers + compatibility.py runtime contract checks +``` + +V2 calls the shared continuation core. It does not implement a second continuation algorithm. + +## 3. Three execution phases + +### Phase 1 — conditioning preparation + +- Resize first/last images once. +- VAE-encode their keyframe latents once. +- Tokenize and encode each unique Prompt once. +- Clone the externally patched MODEL once and add the Continuum APPLY_MODEL wrapper. + +No H3 diffusion sampling begins until this phase completes. + +### Phase 2 — H3-only loop + +For each chunk: + +1. Create native H3 NestedTensor AV latent. +2. Select 5/22/39-frame Context or Auto. +3. Add one native `video` / `video_audio` reference block. +4. Call the standard ComfyUI guider/sampler path. +5. Let Spectrum create and tear down its own run through its public wrappers. +6. Copy the sampled full AV latent to CPU once. +7. Derive the next V1-compatible State from that CPU chunk. + +Video and Audio VAE decode are deliberately absent from this loop. + +### Phase 3 — deferred decode + +- Decode each CPU chunk after all H3 sampling. +- Immediately move decoded tensors to CPU. +- Remove the overlap. +- Copy images into one preallocated final IMAGE tensor. +- Place audio into cumulative 24fps-derived sample ranges. +- Release each raw chunk before decoding the next. +- Apply exact final duration if enabled. + +## 4. Native H3 continuation representation + +V1 and V2 both represent the previous tail as one reference block: + +```text +kind: video or video_audio +latent: native H3 video latent tail +latent_t/h/w: native geometry +audio_latent: optional native H3 audio latent tail +ref_audio_t: audio latent length +_h3cj_* metadata: timeline placement contract +``` + +The outer APPLY_MODEL wrapper receives the payload built by ComfyUI, moves only marked Continuum reference rows onto the target timeline, and normalizes keyframe/reference latent lists. + +## 5. Accelerator composition + +### SageAttention + +Continuum never replaces the attention kernel. Existing attention overrides remain intact. + +### Sol-Attn + +Sol-Attn registers H3 target spans by the identity of `PackedLayout.position_ids`. Continuum modifies this tensor in place and never replaces it. Sol conditioning sink and Morton tracking therefore retain their registration. + +### Spectrum + +Spectrum remains an external OUTER_SAMPLE / PREDICT_NOISE / SAMPLER_SAMPLE / DIFFUSION_MODEL wrapper stack. V2 calls the normal ComfyUI `guider.sample()` once per chunk. Spectrum starts and ends an independent runtime run for each call. Continuum imports no Spectrum private functions. + +The recommended model chain is: + +```text +UNET → Sage → Sol → LoRA → Spectrum → V2 Sampler +``` + +V2 then clones this ModelPatcher once; Spectrum's ON_CLONE callback creates an independent runtime binding on the clone. + +## 6. State and Session + +### V1 State + +Contains only the maximum retained continuation tail and temporal metadata. It can start V2 through `initial_state`. + +### V2 Session + +Contains all accepted raw chunk AV latents plus: + +- Prompt and SHA-256 Prompt hash +- per-chunk Seed +- Continuum Plan +- selected Context frames and motion score +- model/accelerator fingerprint +- branch parent/session identifiers + +Tensors remain on CPU. Save Session writes safetensors and a JSON mirror atomically. The authoritative metadata is embedded in safetensors. + +## 7. Reroll semantics + +- `reroll_from_chunk=0`: reuse every matching accepted prefix. +- `reroll_from_chunk=N`: preserve chunks before N and regenerate N onward. +- A Prompt mismatch automatically stops reuse before the changed chunk. +- A first-frame fingerprint mismatch fails explicitly. +- A model/accelerator fingerprint mismatch is reported but accepted chunks remain fixed, because their latent output is already approved. +- The new output creates a child Session; the input Session is not mutated. + +## 8. Audio timeline + +H3 video is 24fps and audio latent is 40Hz. Decoded PCM is placed using cumulative integer boundaries: + +```text +start_sample = round(cumulative_frames / 24 * sample_rate) +stop_sample = round(next_cumulative_frames / 24 * sample_rate) +``` + +This avoids the one-sample drift that can occur when every 119-frame segment is rounded independently. + +## 9. Memory model + +Full accepted AV latents are comparatively small and stored on CPU. The final decoded IMAGE is large, so V2: + +- estimates final float32 IMAGE + one raw chunk + headroom before sampling; +- refuses unsafe jobs before expensive H3 work; +- allocates the final IMAGE exactly once; +- copies each decoded segment into its final range; +- avoids progressive whole-sequence concatenation. + +## 10. Compatibility policy + +The package checks: + +- native `MiniMaxH3Model` class and required attributes; +- `PackedLayout` capability, including transparent `*args/**kwargs` wrappers; +- H3 temporal grid `(1,4,4,4,4)`; +- APPLY_MODEL and DIFFUSION_MODEL wrapper APIs; +- NestedTensor support; +- real PackedLayout row placement through the installer self-test. + +Marked Continuum payload failures never silently fall back to stock reference placement, because that could render a plausible but semantically wrong join. diff --git a/docs/FAILURE_SEMANTICS_BASELINE.md b/docs/FAILURE_SEMANTICS_BASELINE.md new file mode 100644 index 0000000..3be74a8 --- /dev/null +++ b/docs/FAILURE_SEMANTICS_BASELINE.md @@ -0,0 +1,93 @@ +# R0 Failure Semantics Baseline + +Date: 2026-09-14 +Scope: current V3.8 Python-exception behavior +Storage root: pytest `tmp_path` only + +## Isolation contract + +すべてのfailure injectionは一時Run Storage rootで実施した。 + +```text +real h3_continuum storage touched: NO +existing Take / revision touched: NO +user backend 8188 touched: NO +GPU used: NO +public workflow changed: NO +``` + +このbaselineは、catch可能なPython exceptionが`RunStorageController.__exit__()`へ到達する場合を記録する。OS強制終了、電源断、process killの挙動を成功扱いしない。 + +## Results + +| Injection point | Committed records after cleanup | Manifest status | Canonical pointer | Current observable behavior | +|---|---:|---|---|---| +| before Sampling | 0 | `interrupted` | none | errorを記録し、review pause metadataを除去 | +| after Sampling / before chunk commit | 0 | `interrupted` | none | sampled resultは永続化されない | +| after first entry of terminal physical group | 1 | `interrupted` | none | terminal pairの片側raw/recordがinterrupted revision内に見える | +| before manifest update (one-shot exception) | 1 | `interrupted` | none | in-memory recordを`__exit__()`が再度manifestへ書き、interruptedとして回収 | +| after manifest update | 1 | `interrupted` | none | committed recordを保持してinterrupted化 | +| before output projection | 1 | `interrupted` | none | recordを保持してinterrupted化 | +| during output projection | 1 | `interrupted` | none | recordを保持してinterrupted化 | +| before storage finalize | 1 | `interrupted` | none | in-progress recordを保持してinterrupted化 | +| before canonical pointer update (first `_write_project`) | 1 | `interrupted` | none | finalize途中のmanifestをexception cleanupがinterruptedへ戻し、2回目のproject writeはcanonicalを設定しない | + +Assertions are implemented in: + +```text +tests/test_runtime_redesign_r0_characterization.py +``` + +## Current write ordering + +### `commit_chunk()` + +```text +validate one logical entry +→ write temporary safetensors +→ fsync temporary +→ os.replace to chunk_000N.safetensors +→ fsync chunks directory +→ construct one record +→ replace manifest chunks in memory +→ write manifest +``` + +この処理はphysical group単位ではなくlogical entry単位である。terminal merged pairでも、1件目commitと2件目commitの間に例外が起きる窓がある。 + +### `finalize()` + +```text +validate committed prefix and Session length +→ set status / review metadata +→ write manifest +→ build branch provenance +→ write manifest again +→ write project with set_canonical=True +``` + +Python exceptionが上位のstorage scopeへ戻れば、`__exit__()`はstatusを`interrupted`へ変更し、canonical指定なしでprojectを再書込する。 + +## Process-crash limitations + +このR0 testは、次を保証しない。 + +- `os.replace(raw)`直後、manifest更新前にprocessが消失した場合のorphan raw回収 +- 既存`chunk_000N.safetensors`を上書き中にprocessが消失した場合の旧manifest/raw整合性 +- terminal pairの両entryを不可分にpublishすること +- manifest provenance書込後、project canonical更新前にprocessが消失した場合の再起動時selection + +これらはR5のprocess-crash consistency対象である。R0でwrite orderやschemaを変更していない。 + +## R5-approved delta boundary + +将来R5で許可される差分はV9どおり次に限定する。 + +- physical group全entryを先にvalidate +- immutable transaction raw filenames +- raw durabilityとSHA verification後にmanifestを1回でswitch +- `commit_group()`はcanonical pointerを更新しない +- canonical pointerはsuccessful finalize後だけ更新 + +Sampling結果、Session/State/Assembly、Review user-visible semantics、Run Storage v3 readabilityは変更しない。 + diff --git a/docs/FAILURE_SEMANTICS_DELTA_R5.md b/docs/FAILURE_SEMANTICS_DELTA_R5.md new file mode 100644 index 0000000..2074e51 --- /dev/null +++ b/docs/FAILURE_SEMANTICS_DELTA_R5.md @@ -0,0 +1,90 @@ +# Failure Semantics Delta — R5 + +Date: 2026-09-14 +Status: R5 Design Delta PASS / Implementation PASS / Validation PASS +Baseline: authoritative ComfyUI_W dirty tree, with R0–R4 PASS before R5 + +## Approved differences + +R5 may change only these two externally observable failure details. + +### D1: terminal physical-group manifest visibility + +Previously, the first logical entry of a terminal merged pair could become +visible in an interrupted manifest before the second entry was committed. +R5 publishes all records of that physical group in one manifest replacement. +An incomplete terminal group is never manifest-visible. + +After all raw files have been written and verified, catchable Python exceptions +may recover the entire terminal group as `interrupted`. They may not recover +only one logical half. + +### D2: immutable transaction raw names and orphan layout + +Legacy raw name: + +```text +chunks/chunk_000N.safetensors +``` + +R5 raw name: + +```text +chunks/rev--txn--chunk-000N.safetensors +``` + +The reader continues to accept the legacy filename stored in existing v3 +manifests. A retry uses a new transaction filename and never overwrites a raw +file referenced by an older manifest or Take. A process stop may leave an +unreferenced raw or temporary file. R5 does not adopt, promote, or delete that +orphan automatically. + +## Manifest switch definition + +```text +write a complete candidate manifest to a unique temporary file +→ flush and fsync the temporary file +→ atomically replace the active manifest +→ run the available directory durability helper +``` + +The completed atomic replacement is the **manifest visibility switch**. The +subsequent directory helper completion is the **manifest durability +completion**. Process-stop tests observe the state before replacement, +immediately after replacement, and after the directory helper separately. + +On Windows, atomic replacement and media-level durability are not treated as +equivalent. The current Windows directory helper is a no-op; validation records +that limitation and does not claim power-loss or OS-crash durability. + +## Behavior that remains unchanged + +- non-terminal accepted-prefix meaning and resume position +- manifest status and Review/Take behavior +- canonical eligibility, fallback selection, and pointer meaning +- Sampling, Projection, and finalize failure semantics +- accepted chunk tensors, seeds, prompts, Session, State, and Assembly data +- Run Storage v3 schema readability +- Storage OFF filesystem-I/O behavior + +`commit_group()` writes physical-group raw files and publishes their records. +It never updates the canonical project pointer. Canonical promotion remains in +the existing successful-finalize path. + +## Process-stop acceptance + +- Before manifest replacement, a fresh reader sees the previous complete + manifest. Any new raw files are unreferenced orphans. +- After manifest replacement, every raw referenced by the new manifest exists, + has the recorded size, and matches the recorded SHA-256. +- At the replacement boundary, the active file is a complete old manifest or a + complete new manifest. A partial JSON file and a half terminal group are + invalid. +- A forced process stop does not run `__exit__`; it is not required to change + status to `interrupted`. +- Existing canonical fallback selection is preserved even in windows where the + project pointer has not yet been replaced. + +The test fixture covers raw-write, raw-verification, manifest-temporary, +manifest-replace, directory-helper, finalize-manifest, and project-pointer +boundaries with a separately terminated worker and a fresh reader. diff --git a/docs/FUTURE_ROADMAP.md b/docs/FUTURE_ROADMAP.md new file mode 100644 index 0000000..95e1f64 --- /dev/null +++ b/docs/FUTURE_ROADMAP.md @@ -0,0 +1,16 @@ +# Future roadmap after V2 + +V2 is the integrated production path. Future changes must preserve V1/V2 schemas and remain optional. + +Possible additions: + +- streaming decode/save that does not retain the complete final IMAGE tensor; +- interactive per-chunk approval UI with a frontend timeline; +- persisted multiple child branches and branch comparison; +- richer identity diagnostics using optional external encoders; +- adaptive Spectrum seam policy only if Spectrum exposes a stable public hook; +- reference-video / Ref2VA identity-memory mode; +- multi-resolution continuation through explicit regeneration, never silent resize; +- batch-independent server/API orchestration. + +These are not promises for a specific release. The current V2 avoids private accelerator APIs so updates remain maintainable. diff --git a/docs/ISSUE31_SAMPLING_SIGNATURES.md b/docs/ISSUE31_SAMPLING_SIGNATURES.md new file mode 100644 index 0000000..3dd1ba2 --- /dev/null +++ b/docs/ISSUE31_SAMPLING_SIGNATURES.md @@ -0,0 +1,66 @@ +# Issue #31 — Attention and Spectrum reuse signatures (3.9.5) + +## 3.9.10 — Spectrum audit portability + +The repair treats source CRLF and LF line endings as equivalent +in both the provider-file audit and provider module closures. All other source +bytes and the exact supported Python file sets remain checked. The canonical +source hashes are unchanged, so existing valid Python 3.13 signatures retain +their identities. + +Separate `spectrum_provider_v1_py312.json` and +`spectrum_provider_v029_py312.json` profiles identify the same audited source sets +under Python 3.12. They were generated and checked with Python 3.12.10; the +original profiles remain byte-identical for Python 3.13.12. Other Python runtimes, +altered implementations, foreign closures and unknown runtime fields remain +non-reusable. The adapter does not execute provider functions to approve reuse. + +This repairs Spectrum observation, not every Core/Kitchen/KJ combination. +Independent Attention/provider checks must also pass. In particular, an older +Core/Kitchen environment can still fail its own Attention observation even when +Spectrum is recognized. Do not upgrade Python or replace Attention solely to +make an old Take reusable. Keep old history and start a new Run after updating. +No new GPU or reporter-exact acceptance is claimed for this repair. + +The standard CPU/JS gate passed **2,096 tests / 2 skips / 2 known expected +Inspector failures**, with compile and offline registration checks passing. +Separate-process storage probes passed 16 Python 3.13 PyTorch/Kitchen cases and +eight Python 3.12 Spectrum/default-model cases; actual Take read-only JavaScript +matched all 24. The six previously valid full Sampling Contracts stayed exact. +These CPU checks execute no Attention kernel or H3 checkpoint inference. + +### 日本語 — 3.9.10の局所修正 + +Spectrum監査でCRLF/LFの改行差を同一として扱い、対応済みソースにPython 3.12用の +固定プロファイルを追加します。既存3.13の正常な識別値と旧Takeは維持します。 +未知のコード・追加状態の拒否は継続します。Spectrumが識別可能でも、Core/Kitchenなど +別の観測に失敗する環境は再利用不可のままです。適用後は新規Runで確認してください。 +旧履歴を利用可能にする目的でPythonやAttentionを変更する必要はありません。 + +## 3.9.8 Preview Override Phase A (2026-10-16) + +The new adapter is restricted to the KJ source profile in `preview_override_provider_v1.json`: Preview Override SHA-256 `d07c8e5ff7d7abd3e31acefe2cf29f618b5da5341b3b434e72ec2ac0953230b0`, audited sibling helpers and Core preview/dispatcher files. The complete profile digest is part of the persistent identity, so later dependency-profile upgrades cannot retain the same identity accidentally. It compares loaded functions, nested code, defaults and supported classes with the pinned source; it never imports KJNodes, executes a preview callback, loads a VAE or initializes CUDA during signature observation. Source line endings are normalized to LF for provider auditing. + +After that audit, the exact builtin `frames` list is treated as execution-local display data. Max resolution, node ID, JPEG quality, suppression, frame count, FPS and supported Core preview settings remain part of the identity. Core's cross-key insertion order and each wrapper list's order are preserved explicitly, because the generic dictionary observer sorts keys. The ordinary non-KJ signature path is unchanged. + +Phase A requires `vae is None`, `audio_vae is None`, KJ Tiny VAE `none`, and Core NoPreviews/Auto/Latent2RGB. Unknown files, replaced code/defaults/helpers, extra wrapper state, foreign classes and approximate-VAE previews keep reuse disabled; they are not promoted by an empty frame list. This profile does not claim support for every KJNodes 1.5.2 checkout or the different upstream Preview Override source. No general `frames` exclusion is added. + +Existing normal non-KJ Sampling v6 identities remain exact. Earlier safe KJ identities may differ and cannot be assumed reusable; earlier unsafe/v5 histories remain preserved and read-only. Start a new Run after updating; old KJ history is not migrated. Full real-Core CPU gate: 1,995 passed / 3 skipped / 2 existing Inspector expected failures / zero unexpected failures and CUDA initialization attempts; the 48 dedicated KJ tests passed without skips. Normal ComfyUI GPU acceptance passed four Core-API Queues: PyTorch and Comfy Kitchen INT8 + the audited local Spectrum implementation, each with separate-Queue Chunk 1 to 2 continuation and held preview images. Both second Queues reused one chunk and generated one; original Chunk 1 bytes and contracts were unchanged. All four video/audio outputs fully decoded. See docs/KJ_PREVIEW_OVERRIDE_V398.md for exact support and limits. Product files used in this targeted GPU gate are byte-identical to the 3.9.8 repair. Browser Run-button behavior, subjective preview quality, reporter-exact reproduction and per-kernel return tracing are not newly accepted. + +## Previous 3.9.5 Attention / Spectrum repair + +With Model Attention Backend between a LoRA Loader and Spectrum, cyclic Attention closures and a slots-based SpectrumH3Binding could be unobservable. Generation still succeeded, but Sampling v6 correctly refused reuse with "Saved history is read-only". This update identifies the reviewed runtime implementations rather than forcing resume_safe. + +Core PyTorch and Comfy Kitchen INT8 overrides are identified by their loaded functions, defaults, closures, both container paths and numerical/provider state. Spectrum identity includes all 27 effective immutable config fields and audited source/loaded-function contracts. Reviewed history, counters and statistics do not change persistent identity. Unknown sources, foreign module closures, extra runtime state and altered implementations remain non-reusable. + +Two explicit Spectrum source profiles are included: upstream v0.2.29 at commit 806fe498d574a726fecb8818917f00a4b993fd77 (55 provider files), and the prior audited 44-file implementation. Versions/class names alone do not enable reuse. The loaded-function contract is environment-specific: validation uses Python 3.13.12, ComfyUI Core 0.40.0 and comfy-kitchen 0.2.37; other source/runtime combinations may remain read-only. BSA/Flow/Untwist and KJ/Sage patch-node combinations are outside this targeted acceptance. + +Update the custom node and restart ComfyUI. Start a new Run. Existing unsafe/v5 Runs are preserved and are not repaired or promoted; normal unaffected Sampling v6 identities remain compatible. Sampling v6, Graph v4, Run Storage v3, public IDs/socket order and shipped workflow bytes stay unchanged. + +The signature gate passed 152 CPU tests with one original-profile-only skip and no failures; CUDA initialization attempts were zero. Full standard validation passed 1,893 tests with 2 skips and 2 existing Inspector expected failures (zero unexpected failures and CUDA initialization attempts). Public upstream 0.2.29 PyTorch/Spectrum and Kitchen/Spectrum new-Run GPU gates both passed separate-Queue Chunk 1→2 continuation, prefix/contract preservation and complete 512x512 / 24fps / 10-second AV decoding. Actual CUDA backend returns were observed. Existing Inspector reconfigure expected failures, browser acceptance, visual/audio quality and benchmarks remain separate. + +## 日本語 + +Model Attention BackendとSpectrumを併用したReview継続で、Attentionの循環参照とslots付きBindingを観測できず、生成後のTakeが再利用不可になる問題へ対応します。全27設定値・実ソース・読み込まれた関数を識別し、実行中の履歴や統計値は永続署名へ含めません。resume_safeを強制せず、未知の実装や追加状態は安全側で拒否します。 + +公式Spectrum v0.2.29の固定コミットと、以前に監査した44ファイルの実装を別プロファイルとして扱います。別のSpectrum版、別のPython/Core実装、BSA/Flow/UntwistやKJパッチの組合せ全般を保証するものではありません。更新・再起動後は新しいRunを作成してください。以前のunsafe履歴を自動修復せず、既存TakeとWorkflowは保持します。 diff --git a/docs/KJ_PREVIEW_OVERRIDE_V398.md b/docs/KJ_PREVIEW_OVERRIDE_V398.md new file mode 100644 index 0000000..4a70035 --- /dev/null +++ b/docs/KJ_PREVIEW_OVERRIDE_V398.md @@ -0,0 +1,95 @@ +# KJ Model Preview Override continuation — 3.9.8 / Phase A + +The audited Preview Override wrapper can keep its transient image list without changing the MODEL sampling identity. The adapter retains fixed settings, source and loaded implementation checks, the complete audit-profile digest, and Core's cross-key/list wrapper execution order. It does not add a generic exception for any field named `frames`. + +## 3.9.9 public-provider repair + +The source checkout additionally accepts the pinned upstream implementation from +[KJNodes commit 17195fc](https://github.com/kijai/ComfyUI-KJNodes/blob/17195fcfa8a05f601b766a0f09c9733284f47416/nodes/preview_override_node.py). +Its LF-normalized Preview Override SHA-256 is +`1916295463eb790d3e8ea2b0dae2b25344549479b92e90a1d238acddb2dde3ab`, +profile `kj-preview-upstream-17195fc-phase-a` in +`preview_override_provider_public_v1.json` (canonical profile SHA-256 +`da128a4324d8e0a63d893501f8d65652d25da60690afaf1447d80f2a2b67617f`). Its sibling and Core hashes are the +same as the original table below. The wrapper implementation is unchanged; the +node factory additionally resolves display IDs through Core's DynamicPrompt. +Both source and loaded-code audits remain required. This does not cover arbitrary +KJNodes versions, future Core changes or preview-VAE configurations. + +The original profile JSON and successful descriptors are preserved. Profile +registration alone cannot invalidate another profile's Takes. KJ-free contracts +are unchanged. Missing or malformed optional public-profile data disables only +that profile. Ordinary generation keeps its existing fallback. + +Review and Render History distinguish v5 history from v6 Takes whose generation +conditions could not be verified. Bounded saved reasons identify unsupported +preview settings or implementations; missing metadata points to Detailed Report. +Reasons come from the selected Take's storage revision and do not affect its +sampling identity or enable disabled actions. + +For an unsupported KJ/Core implementation, bypass **Model Preview Override** in a +workflow copy and test a new Run. For unsupported preview VAE settings, disconnect +only that preview node's optional VAE/Audio VAE and use Tiny VAE `none`; keep the +generation VAEs connected. For Core approximate-VAE preview, use Auto or +Latent2RGB. Correct the configuration before starting a new Run: repeatedly +restarting Chunk 1 with the same unsupported setup will not repair it. +Old unsafe/v5 Takes remain readable and are never promoted or rewritten. + +Full real-Core CPU regression: **2,075 passed, 2 skipped, 2 known Inspector +expected failures, zero unexpected failures**. The 90-case production-JS fixture +also passed with eight real Run Storage project vectors. These checks overlap +the pytest total and must not be added together. Unmodified 3.9.8 fixtures confirm +exact legacy/KJ-free identities and byte-identical saved Takes; all 21 existing +workflow JSON/ZIP files are unchanged. Offline registration retains twelve nodes +and passes native PackedLayout / Fixed 3x5s checks. Recorded CPU drivers attempted +zero CUDA initializations. The skips cover a local-only Spectrum descriptor +fixture on upstream Spectrum and an unavailable historical V3.5 workflow. + +The additional public profile passed two targeted H3 GPU Queues with cached +preview images and unchanged Chunk 1. Normal-ComfyUI browser Continue passed +separately with the retained original local profile. See +[3.9.9 validation and limits](RELEASE_NOTES_V399.md). The historical v3.9.8 tag +and assets retain their original local-only profile. + +## Exact supported profile + +The published v3.9.8 package accepts only the following explicitly audited source combination. KJNodes reports 1.5.2, but that version alone does not identify the supported implementation: this Preview Override file differs from upstream. The 3.9.9 extension is described above. Files are LF-normalized before source hashing. An unsupported provider remains non-reusable rather than blocking ordinary generation. + +| Source | SHA-256 | +| --- | --- | +| KJ `nodes/preview_override_node.py` | `d07c8e5ff7d7abd3e31acefe2cf29f618b5da5341b3b434e72ec2ac0953230b0` | +| KJ `nodes/tiny_vae.py` | `61bf4daa634ef6bb901ff8630071caed12e733d17e7c4ba50082dc0aa5506591` | +| KJ `nodes/ltxv_nodes.py` | `11286583c56e653dbd57f5e8eda6902c84b759f16b359b3bf0c895b43e6b5364` | +| Core `latent_preview.py` | `b323be881ade3388254a72f5463e329f6d0a263b0078b69ae4256eb4e22dbba2` | +| Core `comfy/patcher_extension.py` | `1c7b3b855278d0b95d1c1a705fa0b3c91a265f320e2f97bdeb9d15469d6d4ac7` | + +Profile: `kj-preview-1.5.2-local-d07c8e5-phase-a`. Canonical full-profile SHA-256: `d146bd89895fef5a8d576134d519e63bbf4ac1c35188f70b4de4e4ebfb498fc7`. The provider JSON is shipped with the runtime adapter. + +Preview Override's optional VAE and Audio VAE must be disconnected, Tiny VAE must be exactly `none`, and Core preview must be NoPreviews, Auto or Latent2RGB. A generation Video/Audio VAE is still used normally; the restriction concerns the preview node's optional connections. Core approximate-VAE/TAESD, altered helpers, unknown wrappers, extra state and other providers are outside this profile. Source alone is insufficient: loaded functions, nested code, defaults, classes and selected bindings must match the audited implementation. + +## Update and history protection + +Restart ComfyUI, refresh the browser and create a new Run. Do not delete or rewrite old Runs/Takes. Earlier safe KJ identities may differ; old unsafe/v5 history is not repaired or promoted. Ordinary non-KJ Sampling v6 signatures remain exact. Sampling v6, Graph v4, Run Storage v3, State/Session formats, public node inputs and official workflow files are unchanged. + +## Validation + +Full real-Core CPU validation: 1,995 passed, 3 skipped, 2 existing Inspector expected failures, zero unexpected failures or CUDA initialization attempts. The dedicated real-provider tests passed all 48 cases, including 0/1/multiple held frames, reconstruction, fixed-setting changes, wrapper ordering, source/code/default/helper changes, VAE exclusion, Attention/Spectrum combinations, prefix preservation and old unsafe history. Offline registration retains twelve nodes. + +The initial acceptance used the same repair product bytes. A subsequent recheck installed all 464 files from the published v3.9.8 tag (`58fbd855`), restarted normal ComfyUI, verified the live 3.9.8 startup banner and saved adapter/profile identity, and passed four new-Run GPU Queues: + +| New Run | First Queue | Separate second Queue | +| --- | --- | --- | +| Model Attention Backend PyTorch + KJ Preview Override | Generate Chunk 1; 8 held images | Reuse Chunk 1, generate Chunk 2; 16 held images | +| Model Attention Backend Comfy Kitchen INT8 + audited local Spectrum + KJ Preview Override | Generate Chunk 1; 8 held images | Reuse Chunk 1, generate Chunk 2; 16 held images | + +In both second Queues the KJ MODEL node was cached, the Sampler executed, the stored contract/revision stayed equal, and original Chunk 1 latent bytes were unchanged. All four video/audio outputs fully decoded (first Queue 5 seconds / 120 frames; second Queue 10 seconds / 240 frames, 512x512, 24fps). Saved contracts recorded Sampling v6, resume_safe=true and the pinned adapter/profile identity. Runtime source hashes and stored descriptors verified the repair was actually loaded. + +Conditions: RTX 5060 Ti 16 GB, Python 3.13.12, PyTorch 2.14.1+cu130, Core 0.41.0, frontend 1.55.17, comfy-kitchen 0.2.37; 8-step Turbo LoRA, Euler/simple, 2x5-second chunks, Review Each Chunk / Continue, saving and Auto Resume on. Preview Override: resolution 512, quality 80, one frame, 12fps, suppress default on, VAE/Audio VAE absent and Tiny VAE none; Core Auto. Spectrum used the previously audited local implementation with offline smoothing/replay, system-RAM history/archive and model-aware off. This does not newly validate upstream Spectrum 0.2.29 combined with KJ, other wrapper orders, Sage patch nodes, or all preview configurations. + +These were controlled submissions through the normal Core API. Browser Run-button/save-reload behavior, reporter-exact reproduction, per-kernel CUDA return tracing, subjective audiovisual/preview quality and performance benchmarks are not newly accepted. Held PNGs were 32x32, so the image counts do not establish high-resolution preview quality. The two known Inspector reconfigure expected failures remain. Existing workflow files and prior Takes were hash-protected; no history migration occurred. GitHub Release/package verification, Registry upload/moderation and real Manager installation remain distinct publication checks. + +## 日本語 + +監査済みKJ Preview Overrideの一時画像だけを永続署名から分離し、固定設定・実装識別・ラッパー順序を保持します。KJNodes 1.5.2全体ではなく、上表のソースと読み込み済み関数を照合できる実装だけが対象です。VAE/Audio VAE未接続・Tiny VAE none・Core NoPreviews/Auto/Latent2RGBが条件です。 + +更新後は再起動し、新Runを使用してください。旧Take・Workflowの削除や移行は不要で、旧unsafe履歴は再利用不可のまま保持します。CPU 1,995成功(専用48件を含む)、PyTorchおよびKitchen+ローカルSpectrumの各2 Queueで画像保持後のChunk 1→2継続、元Chunk 1保持、4本の映像・音声全デコードを確認しました。初期の同一製品コードによる受入後、公開タグv3.9.8(58fbd855)を実環境へ反映・再起動して、新Runで同じ2構成の4 Queueも再確認済みです。画質・速度・実ブラウザー操作・全KJ構成への合格は主張しません。 diff --git a/docs/LOADER_PERSISTENCE_REPAIR.md b/docs/LOADER_PERSISTENCE_REPAIR.md new file mode 100644 index 0000000..1d8af55 --- /dev/null +++ b/docs/LOADER_PERSISTENCE_REPAIR.md @@ -0,0 +1,28 @@ +# Loader persistence repair — 2026-09-21 + +Base: `3d7c2fd6928a545b12b84968b09dc455e7f9da65`. + +Image keeps its Core-backed implementation and Enable convenience toggle. The mode descriptor lookup walks the prototype chain; a tracked Core accessor is called rather than shadowed by a detached closure. Draw/serialize/configure functions and widget collections are not replaced. The former drawing swap could truncate a stable Core widget view, separately from Issue #23's mode-store divergence. New regression models both failures. + +The current paired official V3.8X2 workflows replace H3EasyLoadAudio with Core LoadAudio and H3ContinuumLoadVideo with Core LoadVideo + H3ContinuumVideoAdapter. Core titles, original media names, disabled input branches, video 24fps setting, and all unrelated node values and original links are preserved. Video/audio outputs remain wired through the old output node ID. Core LoadVideo and the converter are kept in the video input group. Disable the converter or entire group, not only the Core input while its adapter remains active. + +Legacy H3EasyLoadAudio/H3ContinuumLoadVideo remain registered/deprecated. Their old backend input contracts remain intact. The new converter has no Enable, file selector, upload handler or custom JavaScript. It uses the established _resample_video_frames() and passes existing audio through without resampling. Source rate 24 is a frame-object no-op; 30/29.97 to 24 uses the existing duration-preserving selection. This does not claim VFR timestamp-aware conversion or motion interpolation. + +No user Windows worktree, installed runtime, GPU, queue, saved workflow, Run Storage or Takes were accessed. A Core 0.36.0/frontend 1.53.6 live-browser save/reopen, A->B->A, direct/right-click Bypass and upload check remains pending. The fixed H3情報チェック handoff is PENDING: cross-task transport is unavailable in this session. No Release/tag or Registry publication. + +Files already saved without their selection/OFF state require one explicit user reselection; never reconstruct them from defaults. Upstream graph replacement may correctly create a new Run Storage revision, so retain an old workflow for resuming its saved run. + +## Validation result + +- Original-code regression: 6 cases; 3 expected failures and 0 collection/runtime errors. +- Focused CPU/Node suite: 46 passed, zero failures/errors/skips. +- JavaScript module syntax, changed Python compilation, workflow JSON/ZIP parity, endpoint consistency and git diff whitespace checks passed. +- Official workflow structural guard preserved every non-loader node, each original link, and non-layout workflow metadata. +- Scope: isolated GitHub Actions CPU runner, not Windows deployment or a real-browser/GPU test. + +## Loader repair full-suite follow-up (2026-09-21) + +- Main repair `f8eb40d8b44e54386bcf3c5197c48f0f71daf6df` passed the 46 focused tests. Its first full CI returned 1376 passed, 3 skipped, and one stale pre-migration workflow fingerprint failure in `test_v38x2_decode_cache_integration.py`. +- Updated only that expected fingerprint to the user-approved migrated workflow (1B2212B7511FAC0B63DFA6E9F006CD8644C72D08C6945AB0E4ED0E722C97A730); all Decode Cache route and output assertions remain. No runtime or workflow bytes changed in this follow-up. +- Full isolated CPU/Node suite: 1377 passed, 3 skipped, zero failures/errors. Actions run 35558988542. Source was snapshotted with `tools/snapshot.ps1` before edits; source/Registry manifest hashes verified. +- Windows deployment, live-browser checks and GPU generation were not performed; fixed H3情報チェック handoff remains pending. No Release/tag/Registry publication. diff --git a/docs/N2B_BRANCH_PROVENANCE_CONTRACT_V1.md b/docs/N2B_BRANCH_PROVENANCE_CONTRACT_V1.md new file mode 100644 index 0000000..02abdc4 --- /dev/null +++ b/docs/N2B_BRANCH_PROVENANCE_CONTRACT_V1.md @@ -0,0 +1,69 @@ +# Branch Provenance Contract v1 + +Status: V3.8 N2b implementation contract. + +## Purpose + +Branch Provenance records which immutable physical-group revision is active at +each point in a Run Storage chain. It allows a user to select an older Take and +generate only the physical groups after that Take without deleting the former +branch or confusing equal nonces with equal ancestry. + +Take labels are presentation only. Internal identity is always `revision_id`. + +## Physical-group revision + +Every completed physical group records: + +- `revision_id`: immutable SHA-256-derived identity; +- `parent_revision_id`: the immediately preceding physical-group revision; +- `group`: one-based `start`, `end`, and `physical_group` values; +- `variation_nonce` and `seed_identity`; +- `storage_revision_id` and `generation_contract_sha256`; +- `lineage_sha256` and immutable chunk record identities; +- `created_utc` plus deterministic `revision_order`; +- optional `branch_cut` metadata. + +Terminal Merge `[N-1,N]` is one physical group. A chain that contains only one +logical half is invalid. + +## Canonical chain + +`project.json` has exactly one `canonical_chain`, one +`canonical_head_revision_id`, and an `active_revisions` map. Browsing Takes does +not change them. `Use This Take` or `Continue From Here`, followed by the normal +Queue action, creates a finalized Schema v3 storage revision and atomically +switches the canonical chain. + +`Continue From Here` keeps the selected prefix exactly and starts a new existing +sampling nonce boundary at the next physical group. The Sampling Contract itself +remains version 5 and is not extended with provenance fields. + +## Run Storage Schema v3 + +Schema v3 separates `revision_id` (storage execution identity) from +`sampling_revision_id` (the unchanged Sampling Contract identity). A finalized +manifest contains Branch Provenance v1 metadata with its complete active chain, +branch selection/cut when applicable, and a monotonic `canonical_sequence`. + +The write order is: + +1. save and fsync new chunk payloads; +2. validate all logical records in the completed physical group; +3. atomically write the finalized manifest and its active chain; +4. atomically replace `project.json` with the new canonical head. + +A failure before step 4 leaves the preceding canonical project head intact. +If `project.json` is corrupt, the highest validated `canonical_sequence` in a +finalized manifest is the recovery source. In-progress and interrupted revisions +are not Render History Take candidates. + +## Schema v2 compatibility + +Schema v2 manifests remain readable and are never rewritten merely by loading +them. Their contiguous records are interpreted in memory as a legacy chain with +one canonical physical-group revision per group. The first Schema v3 save uses a +separate storage revision directory, so the legacy manifest and payloads remain +unchanged. + +No automatic Take garbage collection is part of N2b. diff --git a/docs/R0_RUNTIME_CHARACTERIZATION.md b/docs/R0_RUNTIME_CHARACTERIZATION.md new file mode 100644 index 0000000..f8f726d --- /dev/null +++ b/docs/R0_RUNTIME_CHARACTERIZATION.md @@ -0,0 +1,215 @@ +# Runtime Redesign R0 Characterization + +Date: 2026-09-14 (Asia/Tokyo) +Phase: R0 only +Architecture changes: none + +## 1. Historical baseline + +この文書は2026-09-14時点のR0 CPU characterizationです。現行3.9.1の受入結果ではありません。当時の実装・Workflow・既存変更を保持した状態で調査し、アーキテクチャ変更やGPU生成は行っていません。 + +## 2. Baseline validation + +R0変更前: + +```text +full CPU pytest: 1187 passed, 1 skipped, 1 warning, 62.18 s +frontend_review_queue.cjs: 46 passed +tools/verify_runtime.py: PASS +registered public nodes: 7 +``` + +実ComfyUI Queue、GPU、VAE DecodeはR0の対象外であり未実行。 + +## 3. Continuation source baseline + +### 3.1 Storage OFF + +| Inputs | Accepted prefix | Effective generation start | Continuation state | Current result | +|---|---:|---:|---|---| +| none | 0 | Chunk 1 | none | fresh Full Run | +| compatible explicit Session with 1 accepted chunk | 1 | Chunk 2 | last accepted Session entry | prefix reuse | +| Session + `Regenerate From = 2` | 1 | Chunk 2 | Session Chunk 1 | Chunk 2以降を再生成 | +| incompatible explicit Session | 0 | Chunk 1 | none | Sessionをadvisory rejectしてfresh run | +| initial State only | 0 | logical Chunk 1 | supplied State | generated plan is continuation and increments State `clip_index` | +| compatible Session + valid State | Session prefix | prefix後 | Session last entry | StateはSession検証前に抑止 | +| invalid Session + valid State | 0 | Chunk 1 | none | Stateへfallbackせずfresh run | +| initial State + invalid `Regenerate From > 1` | n/a | n/a | n/a | `SequenceRuntimeError` | + +重要なV9との差分: 現在の`_preserved_prefix()`はSession validation、resolution、duration、identity不一致を例外として外へ出さず、noteを付けて空prefixを返す。このため、V9の「invalid/incompatible Sessionなら既存Sessionエラー」は現行実装の事実ではない。 + +実際の現行結果は次である。 + +```text +Session present + initial State present +→ initial State = None +→ Sessionを検証 +→ Session invalid/incompatibleならSession prefixを0件へ落とす +→ initial Stateには戻らない +→ fresh Chunk 1を生成 +→ reportへ "saved session was ignored; generated a fresh run (...)" +``` + +したがってR1以降で維持すべきbaselineは、ユーザーが別途変更を承認しない限り「no State fallback + advisory fresh run」である。 + +Storage OFF characterizationでは、Run Storage Review policy invocationは0、Run Storage filesystem I/Oは0だった。 + +### 3.2 Run Storage ON + +| Inputs / saved state | Current result | +|---|---| +| explicit Session present | `RunStorageError: Run Storage cannot be combined with an explicit Session`。storage書込前に停止 | +| validated saved prefix + initial State | saved prefixが優先。Stateはcontinuation sourceにならない | +| no saved prefix + initial State | initial Stateを使用 | +| Review Continue | review decisionを作成後、prefix実読込後にreconcileして再度decisionを作成 | + +現在のReview Continue decision-producing pathを隔離すると、pure policy callは2回だった。 + +```text +_resolve_review_contract() +→ resolve_review_execution() # first decision +→ prefix selection/load +→ _reconcile_review_execution_with_reused_prefix() +→ resolve_review_execution() # second decision +``` + +`_validated_review_head()`も同じpure functionをmetadata validationに使用するが、その戻り値はQueue decisionではない。R2の「Decision 1回」は上記decision-producing 2経路を1経路にする課題であり、R0では変更していない。 + +## 4. ContinuationSourceFacts mapping baseline + +この型とprojection関数はR0時点のsource treeには存在しない。R3/R4で新しい意味を作らないため、現行runtime値からのmappingを次で固定する。 + +| selected kind | `accepted_chunks` | `reroll_from_chunk` | `physical_group_boundaries` | +|---|---:|---|---| +| `run_storage` | 実際にraw検証とloadを通過した`best_entries`数 | final branch contractの有効boundary。通常continueは0 | accepted prefix内で完全に成立した現行physical groups | +| `explicit_session` | `_preserved_prefix()`が受理したentries数 | explicit Session経路の実効`reroll_from_chunk` | terminal atomicity適用後、accepted prefix内で完全に成立したgroups | +| `initial_state` | 0 | 0または現行で許可される1 | empty tuple | +| `none` | 0 | 0(plain Full Run baseline) | empty tuple | + +`initial_state`はaccepted raw prefixではない。logical Chunk 1のSamplingにcontinuation stateを供給するだけである。 + +## 5. Current validated-prefix representation + +R0 source treeには`ValidatedPrefix`、`EmptyValidatedPrefix`、`PrefixFacts`、`ContinuationSourceFacts`、`to_planning_facts()`はまだ存在しない。現在は次のmutable valuesへ分散している。 + +- `review_head`: `revision_id`, `status`, `validated_prefix_count`, `review_unit`, `branch_regenerate_from`, `effective_reroll_nonce`, `updated_utc`, `manifest` +- `best_entries`: raw existence、SHA-256、entry contractを通過して実際にloadされたentries +- `best_records`: accepted entriesに対応するmanifest records +- `RunStorageController.reused_count`: `len(best_entries)` +- `RunStorageController.review_execution`: prefix load前decisionを、load後reconcileで置換した値 + +よってV9の`ValidatedPrefix.to_planning_facts() -> PrefixFacts`は「現行契約」ではなく、R2で導入する将来契約である。R0でその型を追加するとR1/R2実装へ先行するため、追加していない。R2で導入する場合も戻り型を`ContinuationSourceFacts`へ流用しない。 + +## 6. ReviewUnit inventory + +Productionの`ReviewUnit`定義は1種類だけである。 + +```text +v3/review_control.py + class ReviewUnit(start, end, physical_group) +``` + +Production usage: + +- `v3/review_control.py`: unit作成、normalization、pause metadata、Review/Take policy +- `run_storage.py`: import、review group marker、pause metadata validation、finalize validation +- `v3/nodes.py`: `review_execution`からprojection rangeを読む。別`ReviewUnit`型は定義しない + +R0検索では別の`class ReviewUnit`は存在しなかった。 + +## 7. Required behavior coverage map + +新しいR0 testは`tests/test_runtime_redesign_r0_characterization.py`に隔離した。既存fixtureと合わせた代表coverageは次の通り。 + +| Contract | Characterization coverage | +|---|---| +| Review Q1 → next → complete | `test_limit_one_returns_one_new_chunk_and_prefix_assembly`, `test_prefix_five_generates_only_chunk_six_and_completes`, L0/L1 storage tests | +| Regenerate Current | N0-N5 and terminal regenerate tests in `test_v38_review_run_storage.py` | +| Finish Remaining | `test_n5_finish_remaining_keeps_current_branch_nonce`, `test_finish_remaining_complete_records_the_final_generated_group` | +| Use This Take / Continue From Here | N2B take-selection, canonical, and continue tests | +| Back to Settings compatible extension | `test_i2va_two_to_three_reuses_both_canonical_prefix_chunks` | +| terminal physical group | terminal atomic tests in execution-cap and Run Storage suites | +| Session/State precedence | new R0 characterization tests | +| Balanced 22 / Strong 39 | new parametrized R0 characterization test | +| Reference Image / Audio and Refine Context | `test_audit_boundary_repairs.py`, `test_v35_context_sampler.py` | +| Video Reference | `test_p46a_reference_video_memory.py` and sequence contract tests | +| Timeline Video | `test_timeline_video.py` | +| Driving Audio | `test_v34_driving_audio.py` | +| Second Pass | `test_v35_second_pass_contract.py`, `test_v35_second_pass_sampling.py`, selective node tests | +| external wrappers | graph-contract Turbo/Sage tests and `test_model_patch.py` | + +## 8. Hardening signature baseline + +`hardening.run_sequence_with_hardening()` calls `inspect.signature(base).bind_partial(...)`. The current keyword-only `run_sequence` signature contains 41 parameters, in this order: + +```text +model +clip +video_vae +audio_vae +sampler +sigmas +first_frame +last_frame +prompt_plan +width +height +continuity +base_seed +audio_continuity +exact_total_duration +diagnostics_mode +reroll_from_chunk +reroll_nonce +strict_compatibility +debug +seam_correction +enable_preview +session +initial_state +latent_only +reference_assets +reference_audio_source +reference_audio_vae +driving_audio_source +driving_audio_vae +reference_video_source +timeline_video_source +guide_source +capture_refine_context +memory_attribution +prompt_conditioning_cache +reference_encode_cache +continuation_transport +max_new_physical_groups +_memory_attribution_collector +_diagnostic_continuation_policy +``` + +R4 thin adapterはこの名前・keyword-only形・defaultsを維持する必要がある。 + +## 9. R1 concrete plan (not implemented) + +R1はProjectionだけに限定する。 + +1. 現在`v3/nodes.py`にある`_review_decode_chunk_range()`の判断を、torch/ComfyUI非依存のpure `ProjectionDecision`へ写す。 +2. `full_accepted_prefix`と`current_review_unit(start,end)`だけを表現する。 +3. 現在の`_apply_review_decode_scope()`のTensor/Assembly再構築をOutput Projectorの単一ownerへ移す。 +4. Session、full Refine Context、Run Storage recordsは切り詰めず、表示用Video/Audio latentとAssemblyだけを同じrangeで投影する。 +5. terminal merged unitは既存physical group境界に従い分割しない。 +6. Full Run、Finish Remaining、complete、Use This Takeは現行どおりfull projectionを維持する。 +7. focused parity、full CPU、frontend、manifest、workflow SHA/ZIP parity、public schemaをGateにする。 + +R1ではReview policy、prefix探索、Session/State priority、Sampling、commit/finalizeを変更しない。R1実装は明示承認までHOLD。 + +## 10. R0 disposition + +R0 Characterizationの実装対象はtest/documentationだけで、Production runtimeは変更していない。 + +R1へ進む前の要確認事項: + +1. V9の「invalid/incompatible Sessionは既存エラー」を、実コードの「advisory reject + fresh run」へ訂正するか。 +2. `ValidatedPrefix.to_planning_facts()`を現行契約ではなくR2導入契約として扱うか。 + +この2点を未訂正のままR2/R4へ進むと、現行挙動維持とV9本文を同時に満たせない。 + diff --git a/docs/R6_NODE_FACADE_DELTA.md b/docs/R6_NODE_FACADE_DELTA.md new file mode 100644 index 0000000..a93c3d6 --- /dev/null +++ b/docs/R6_NODE_FACADE_DELTA.md @@ -0,0 +1,41 @@ +# R6 Node Facade Delta + +## Purpose + +R6 reduces only the public V3.8 Sampler method into this fixed boundary: + +```text +public inputs + -> RuntimeRequest + -> existing hardening/runtime adapter + -> RuntimeOutput + -> public tuple +``` + +`v3/node_facade.py` owns the immutable request container and the V3.8 +status-only resolution/diagnostic decoration. It does not import or create a +Coordinator, Planner, Run Storage controller, Sampling Engine, or ReviewUnit. + +## Preserved ownership + +- `v2/sequence.run_sequence` remains the R4 signature-compatible hardening + adapter and keeps its explicit 41-keyword signature. +- `InternalRuntimeCoordinator` remains the sole owner of continuation-source, + Session/State priority, prefix validation, and Runtime execution ordering. +- `execution_planner` remains the sole owner of physical-group and Projection + decisions. +- `run_storage` remains the sole owner of R5 immutable raw transactions and + manifest publication. +- The V3.8 node still resolves its existing Size Source contract before the + request boundary and calls the unchanged V3.7 runtime inheritance path once. + +## Non-goals and compatibility + +No public node ID, socket, widget order, output tuple, Workflow, Sampling, +Review, Run Storage schema, canonical selection, or hardening behavior is +changed. Tensor/model identities are preserved; only keyword containers are +copied into immutable request mappings. Duplicate runtime keywords fail +instead of being silently overridden, matching the prior explicit-keyword +call's fail-closed behavior. + +This phase introduces no temporary adapter, new dependency, or future hook. diff --git a/docs/README_HISTORY.md b/docs/README_HISTORY.md new file mode 100644 index 0000000..1b8bd13 --- /dev/null +++ b/docs/README_HISTORY.md @@ -0,0 +1,905 @@ +# H3 Continuum — README history and implementation reference + +[Current README](../README.md) · [Illustrated user guide](USER_GUIDE.md) · [CHANGELOG](../CHANGELOG.md) + +This reference preserves the release descriptions, earlier screenshots, detailed +validation results and pre-V3.8 implementation notes previously shown in README. +It was extracted from source commit `894869d9` (package 3.9.8). +Statements such as "current on main", historical installation commands, node lists +and old acceptance results retain their original context; they are not current +installation recommendations or broader validation claims. +For current downloads, compatibility and known limits, start with the README. + +- [v3.9.8 and published-tag GPU conditions](#398--kj-model-preview-override-continuation) +- [Older implementation notes](#archived-implementation-notes-pre-v38-not-the-current-guide) +- [Prompt formats and examples](#prompt-formats) +- [Historical resource and validation limits](#limits) + + + + +**Latest package: 3.9.8.** Use the fixed-version GitHub downloads below. Registry publication is verified separately. + +| Version | Release | Custom-node ZIP | SHA-256 checksum | +| --- | --- | --- | --- | +| 3.9.8 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.8) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.8/ComfyUI-H3-Continuum-v3.9.8.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.8/SHA256SUMS.txt) | +| 3.9.7 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.7) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.7/ComfyUI-H3-Continuum-v3.9.7.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.7/SHA256SUMS.txt) | +| 3.9.6 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.6) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.6/ComfyUI-H3-Continuum-v3.9.6.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.6/SHA256SUMS.txt) | +| 3.9.5 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.5) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/ComfyUI-H3-Continuum-v3.9.5.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/SHA256SUMS.txt) | +| 3.9.4 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.4) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/ComfyUI-H3-Continuum-v3.9.4.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/SHA256SUMS.txt) | +| 3.9.3 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.3) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/ComfyUI-H3-Continuum-v3.9.3.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/SHA256SUMS.txt) | +| 3.9.2 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.2) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/ComfyUI-H3-Continuum-v3.9.2.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/SHA256SUMS.txt) | + +## 3.9.8 — KJ Model Preview Override continuation + +An audited KJ Model Preview Override can keep preview images without changing a new Run's sampling identity. Separate Queues can reuse Chunk 1 and generate Chunk 2 while preserving earlier Takes. Support is limited to the explicitly pinned KJ source and Core helpers, **VAE / Audio VAE disconnected, Tiny VAE `none`**, and Core NoPreviews/Auto/Latent2RGB. Other KJ builds remain conservative; version 1.5.2 alone is insufficient. Restart ComfyUI, refresh the browser and start a new Run; do not delete old Takes or workflows. See [compatibility, source hashes and verified scope](KJ_PREVIEW_OVERRIDE_V398.md). + +### Published v3.9.8 GPU recheck — 2026-10-16: PASS + +After restarting the normal ComfyUI installation with the published tag **`v3.9.8` / `58fbd855`**, all 464 installed release files matched the tag. The startup log confirmed 3.9.8, and new saved Sampling v6 contracts contained the KJ adapter and pinned audit-profile digest. Two brand-new Runs passed four Core-API GPU Queues: + +| KJ Preview Override configuration | Separate-Queue Chunk 1 → 2 | +| --- | --- | +| Model Attention Backend: PyTorch, Spectrum OFF | **PASS — 1 reused / 1 generated** | +| Model Attention Backend: Comfy Kitchen INT8 + audited local Spectrum | **PASS — 1 reused / 1 generated** | + +On both second Queues, the KJ MODEL stayed cached with preview images retained (**8 → 16**), while the Sampler executed. The contract/revision and original Chunk 1 latent hashes stayed unchanged; `resume_safe=true`. All four video/audio outputs fully decoded (5 seconds / 120 frames, then 10 seconds / 240 frames, 512×512 at 24fps). Existing workflows and Takes were preserved: 170 protected files, including 71 saved workflow files and 46 pre-existing Run/Take files, retained their hashes. + +**Scope:** only the exact audited KJ/Core profile; the local KJ source differs from upstream, so KJNodes 1.5.2 alone is insufficient. Preview Override VAE / Audio VAE were disconnected, Tiny VAE was `none`, Core preview was Auto. GPU: RTX 5060 Ti 16 GB, 8-step Turbo, Euler/simple, 2×5-second chunks. Spectrum used the audited local offline-replay implementation, system-RAM history/archive and model-aware OFF. **Unverified:** browser Run-button/save-reload behavior, the reporter's exact complete graph, other KJ/Spectrum builds, SolAttn/preview VAE paths, per-kernel return tracing, subjective quality and speed/VRAM benchmarks. Retained PNGs were 32×32, not high-resolution preview-quality evidence. The known Inspector issue remains; old unsafe/v5 Takes are not repaired or promoted. [Issue #31 result](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/31#issuecomment-6209571261). + +## 3.9.7 — External-input regeneration cache fix + +Fix automatic Full Run variations when Chunks or Variation Nonce come from external nodes. Unresolved linked inputs no longer let Core skip an explicitly requested regeneration. Earlier accepted chunks, existing Takes, public inputs and official workflows are preserved. Update the custom nodes, restart ComfyUI and refresh the browser; no workflow changes or history deletion are required. [Release notes](RELEASE_NOTES_V397.md). + +**Published v3.9.7 GPU recheck (2026-10-15): PASS.** After restarting with the published tag (`18df80d`), external Core INT nodes supplied Chunks=2 and Nonce=0. Two completely identical Chunk 2 regeneration prompts advanced automatic Nonce **1 → 2**; both reused Chunk 1 and generated only Chunk 2. Earlier Takes and raw latents were retained. All three outputs (baseline plus two variations) fully decoded with 240 frames and 10-second video/audio. Existing 60 workflows, one settings file, 30 Run/Take files and all 459 release files remained hash-identical. Scope: RTX 5060 Ti 16 GB, 512×512, 24 fps, 2×5 seconds, 8-step Turbo, PyTorch attention, Spectrum OFF, Core API submissions. No CUDA/OOM/execution errors were observed; this does not extend browser, subjective-quality or other-accelerator acceptance. + +The original [3.9.2 GitHub source archive](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/archive/refs/tags/v3.9.2.zip) remains available from its unchanged tag. This custom-node package ZIP is distinct from the official workflow ZIPs. + +{56C36D94-39EC-4484-9495-5648A1F486F3} + + +## 3.9.6 — regeneration selections and automatic variations + +Explicit **Regenerate From** and **Variation Nonce** survive V3.8/V3.9 workflow save/reload. With progress saving enabled, **Full Run + an explicit Chunk + Nonce 0** reaches Run Storage on every Queue, allowing another automatic variation after a completed Take. Earlier accepted chunks and previous Takes are kept. Restart ComfyUI and refresh the browser. For an older unmarked Chunk 1 / nonce 0, select Chunk 1 again and save once. See [operation, compatibility and verified scope](REGENERATE_INTENT_V396.md). + +## 3.9.5 — Issue #31: reusable Attention and Spectrum contracts + +Review Each Chunk can resume new Runs using Core Model Attention Backend with PyTorch or Comfy Kitchen INT8 and an explicitly audited Spectrum binding, including upstream Spectrum v0.2.29. Existing unsafe history remains readable and read-only for reuse; start a new Run after updating. Public node IDs, workflows and Sampling v6 are unchanged. See [supported implementations, update steps and validation limits](ISSUE31_SAMPLING_SIGNATURES.md). + +## 3.9.4 — Registry packaging repair (2026-10-14) + +Include the declared Review restoration guide and exclude repository-only prompt-authoring files from the Registry ZIP. The publishing workflow now builds the actual package and checks its exact file set, CRC and every Registry Manifest hash before uploading. Runtime behavior, twelve public node IDs, workflows and saved Runs/Takes are unchanged. GitHub Releases 3.9.3 and 3.9.2 remain available. + +## 3.9.3 — choose temporary storage (2026-10-13) + +Choose where Continuum writes its temporary decoded frames and Finalize buffer. Open **Settings → H3Continuum → Storage → Manage temporary storage**. Enter an absolute folder on the **ComfyUI server**, or choose a server storage candidate, then validate and save. Leave it blank to keep the existing locations. No workflow changes are required; continue using the V3.9 nodes and your 3.9.1 / 3.9.2 workflows. + +- The selected folder applies to **both Finalize and Decode Cache Helper**. It contains a dedicated `h3-continuum-scratch` directory. Saved Runs/Takes and finished videos keep their existing locations. +- The panel shows current and previous locations, temporary file sizes, free space, and files in use or eligible for deletion. **Delete unused temporary files** protects live outputs, Tensor views, active processes and files whose ownership cannot be verified. Deleting a cache can require decoding and writing it again. +- Save or clean only when the running and pending queues are empty. New files use the new location; existing outputs remain valid at their old locations. A missing selected drive does not silently redirect writes to your original SSD. Shared storage controls are read-only in multi-user mode. +- Mode selection stays in the existing nodes: **Finalize: Auto / RAM / Disk-backed**; **Decode Cache Helper: Auto / RAM / Off**. Auto keeps its conservative 4 GiB final-IMAGE threshold and RAM reserve. Manual RAM now checks available memory before allocation; manual Disk-backed checks free space. These checks cannot guarantee against OOM, changing disk usage or a drive disconnect during execution. No new VRAM storage mode is added. + +**Moving storage changes where writes occur, not their total volume.** With only one physical drive, moving to another folder or partition on that drive does not divert its writes. A separate SSD is usually the faster alternative; an HDD can be used but may slow decoding, assembly and saving. If sufficient CPU RAM is available, RAM modes can avoid these scratch writes, but memory pressure can cause paging or OOM. Turning Decode Cache Off does not turn off Finalize's Disk-backed buffer. File-size displays are logical sizes, not physical disk allocation or cumulative SSD writes. Local filesystems are supported; network shares are outside this feature's supported scope. + +Update the custom node package, restart ComfyUI to load the backend, and hard-refresh the browser. Later changes to this storage setting take effect without a restart. Blank settings retain the ComfyUI temp directory for Finalize and `H3_DECODE_CACHE_TEMP_DIR` (or OS temp) for Decode Cache. A selected folder takes precedence over that cache environment variable. + +The storage changes preserve frame dtype, video/audio assembly and generation identity. Validation and remaining platform limits are recorded in [PACKAGE_VALIDATION.txt](../PACKAGE_VALIDATION.txt). The implemented settings screenshot appears at the top of this README. GitHub Release and Registry publication status are listed above; neither follows automatically from a source update. + +If the panel reports HTTP 404, inspect the ComfyUI startup log. An old process can keep the chosen address/port occupied, so the updated process fails to start while its new JavaScript is already visible. Stop the obsolete instance or choose another port, then start the updated backend. **Refresh** retrieves storage status; it does not restart ComfyUI. Failed reads now show the HTTP result and keep mutations disabled until status is available. + +## 3.9.2 — Issue #30 frontend hotfix (2026-10-13) + +This hotfix reduces unnecessary Review and Reference Images UI work: + +- Skip upstream Review settings snapshots when no comparison baseline exists. +- Notify connected Samplers once per Reference selection, preserving immediate native values and the original widget callback. +- Skip Reference node resizing when its target dimensions already match its actual dimensions. + +Review monitoring with a baseline and fresh Take/Queue checks remain in place. Sampling, saved Runs/Takes, public node IDs and official workflow bytes are unchanged. Keep using the V3.9 nodes. + +The frontend repair passed **1,703 CPU/JavaScript regression tests, 2 skipped**, with **2 explicit expected failures** for the independent Reference Plan Inspector bug after repeated Sampler configuration. Offline Core registration/planning and isolated Chromium checks passed. Real ComfyUI browser checks for this package are recorded in [PACKAGE_VALIDATION.txt](../PACKAGE_VALIDATION.txt); mobile freeze resolution is unconfirmed. Reference table rebuilding and Review snapshots with a baseline remain. No new GPU generation is part of this hotfix's validation. + +After updating, **hard-refresh the browser (Ctrl+F5)** to load the JavaScript. Restart ComfyUI normally to update the displayed package version. A 3.9.1 Run/Take does not require migration for this UI hotfix. Versioned downloads are listed in [GitHub Releases](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases); Registry publication is separate. + +## 3.9.1 — what changed (2026-10-04) + +This patch fixes Second Pass conditioning, Take reuse and Review audio: + +- **Second Pass:** preserve audio-only/mixed keyframes and inherit the verified conditioning and Reference assignments of the actual Review output, including a single middle chunk. +- **Take reuse:** include sampler closures and MODEL CFG/wrappers in compatibility checks to prevent reuse under different generation settings. +- **Review Driving Audio:** cut the source PCM once at the physical group's natural time, with the same start position for Exact ON/OFF. +- **Audio resampling:** prefer ComfyUI Core's standard API, with the legacy fallback for older Core versions. The Reference Encode Cache limitation for direct VAE weight changes is documented below. + +**Updating from 3.9.0:** restart ComfyUI and refresh the browser after updating. Existing v5 Takes and history remain readable and are not deleted, but cannot be reused as prefixes for new v6 generation. Start a new Full Video or choose **Start again from Chunk 1**. Public node IDs, widget order and official workflows remain unchanged; keep using the **V3.9** nodes. + +The repaired code passed **1,696 CPU tests / 1 skipped**, targeted Review Second Pass GPU checks and an integrated generation/reuse check. Browser acceptance remains pending. Optional latent-upscale doubled outlines are tracked separately in [Issue #27](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/27); they are not a required gate for these fixes. This is a main-source update, not a new GitHub Release or Registry publication. + +## 3.9.1 repair details + +Second Pass preserves audio-only keyframes and adapts only the video component of mixed keyframes. Review Driving Audio uses the retained natural frame range of the physical decode group, measured before projection. Exact ON/OFF share that PCM origin. Completely empty Review intervals receive silence for the selected video duration; partially short audio stays short with Exact OFF. + +New generation uses Sampling Contract v6 and Graph Contract v4, including sampler code/closures and MODEL CFG/wrapper settings. Saved v5 Takes, raw tensors and provenance remain readable, but cannot be reused in v6 generation. Start a new Full Video or use Review's **Start again from Chunk 1** to create a new revision. Old Takes remain intact. Unobservable external settings still allow fresh generation with reuse disabled; incompatible explicit Take/Retry requests are not replaced with a different Take. + +Reference/Driving resampling prefers Core's standard API and uses legacy TorchAudio only when that API is absent. Preserved output PCM keeps its original waveform and sample rate. + +**Reference Encode Cache limitation:** direct weight mutation inside the same VAE object is not detected automatically. Call `v3.ref_encode_cache.clear_ref_encode_cache()` afterwards or reload into a new VAE object. Unchanged VAE/input pairs continue to hit the cache. Decode Cache Helper reset affects a separate cache and does not clear Reference Encode Cache. + +This repair stays within V3.9: public node IDs, socket/widget order, official workflows, Run Storage v3 and State/Session formats are retained. LoRA Plan is not included. CPU checks and browser/GPU acceptance are separate. + +V3.9 also projects captured conditioning onto the actual Review output groups before Second Pass. A partial prefix or a single middle chunk can inherit its own conditioning and Reference assignments when the context validates. Full-sequence completion and saved Takes are unchanged; missing or mismatched context keeps the diagnostic prompt-only fallback. + +The repaired runtime passed **1696 CPU tests, 1 skipped**, and three targeted GPU cases at **512×512 / 24 fps**: a two-group partial prefix, Chunk 2 alone, and complete three-group reuse. All groups inherited verified conditioning; saved FLAC matched the corresponding source PCM exactly and existing Takes/raw files were preserved. This validates the targeted repair. Browser acceptance remains pending; optional latent-upscale double outlines are a separate investigation in Issue #27. Historical V3.8 releases and tags remain available unchanged. + +![H3 Continuum V3.9: fixed Reference Image slots, per-chunk assignments, and one Sampler input](images/v39/v39-feature-summary.png) + +*V3.9 is current on `main`: fixed Image 1–9 slots, per-chunk Reference selection, and one Reference Images connection to the V3.9 Sampler.* + +## Getting started video + +[H3 Continuum V3.9 — Getting Started: Chunks, Prompts & Reference Images](https://youtu.be/AKJxBXaiG6Q) (12:26) + +Learn basic First Image setup, chunk settings, Review/Take retries, Timeline prompting, and Reference Images. Includes English narration and subtitles, seven timestamped chapters, version-labeled UI stills, and actual generated examples. + +![H3 Continuum V3.8X2 feature card from its original release](images/v39/v38x2-feature-summary-historical.png) + +*This V3.8X2 card is retained for comparison. Its “CURRENT MAIN” label describes the time it was made; V3.9 is current now. The V3.8 Sampler and V3.8X2 workflow remain available, but their Reference wiring is not automatically converted to V3.9.* + +## External INT Width/Height fix on main (2026-10-01) + +Fixed the visible Width/Height input binding in both V3.8 and V3.9. External INT connections now stay aligned with their displayed controls through workflow save/reload and tab switching. Connected controls are read-only; their displayed numbers are saved Manual reserve values, not a prediction of the upstream node's runtime output. Disconnecting restores local editing. + +To set exact dimensions from Core **Int** nodes, choose **Size Source = Manual** and connect the Width and Height INT inputs. **First Image** still uses the image and Resolution setting; connected dimensions do not override it or automatically change Size Source. If no active First Image is available, the existing Manual fallback applies. Dimension validation and backend input order/defaults are unchanged. + +Validation: **1569 CPU tests passed, 1 skipped**, including 22/22 new frontend regression cases and 60/60 existing Review cases; Chrome save/reload and five tab round trips for both versions. One V3.9 GPU smoke run with external INT values produced exactly **480×640, 96 frames at 24 fps, with 4-second audio** in **107.3 s** (20 sampling steps); the entire saved video decoded successfully. This is a short single-condition check, not long-form or all-model acceptance. Sampling, Run Storage, the earlier Review/cache repairs, official workflow JSONs/ZIPs and historical Releases are unchanged. + +Update Continuum from `main`, then **hard-refresh the browser (Ctrl+F5)** to load the new JavaScript. This fix alone does not require a ComfyUI backend restart; restart normally if the update also brings Python changes or new node definitions. This remains a main-source fix, not a new Release or Registry package. + +## V3.9 performance fix on main (2026-10-01) + +V3.9 could repeat unnecessary prompt-conditioning preparation when **Prompt Format = Fixed** and **no Reference Images are selected**. This repair reuses the existing bounded conditioning cache when its inputs are unchanged. First Image is supported; List/Timeline prompts, selected Reference Images, Reference Audio, Timeline Video and terminal-pair groups keep their existing paths. This is not the Decode Cache Helper and does not skip Sampling. + +In the tested `2 × 5 s`, 480×640 First Image workflow, group preparation fell from about **10.76 s to 0.013 s**. In the post-fix same-session comparison (three measured runs per version, after warm-up), median total time was **273.58 s for V3.8 vs 273.78 s for V3.9**; the ranges overlapped. The earlier roughly 6% gap was no longer observed under these conditions. This does **not** establish a speed fix for every workflow, model or GPU, or a GPU-kernel speedup. + +Update Continuum from `main` and **fully restart ComfyUI** to load the Python changes. Validation: **1558 CPU tests passed, 1 skipped; all 8 GPU generation runs succeeded** on ComfyUI 0.38.0 / RTX 5060 Ti 16 GB. Seeds, chunk/frame contracts, Run Storage, V3.8X2/V3.9 official workflows and ZIPs, and historical Releases are unchanged. This is a main-branch fix, not a new Release or Registry package. + +## Review restoration fix on main (2026-10-01) + +Switching to another workflow and back could hide Review actions even though saved Takes still existed. This frontend repair waits for workflow restoration before reading history and ignores stale results from previous graphs. Real generation-setting changes still disable Review/Take application; restoring the original settings re-enables the actions. + +Update the custom node from `main`, then refresh the browser to load the new JavaScript. Validation: 60/60 frontend regression cases, 108 focused CPU tests, five Chrome tab round trips, and a Chrome Base Seed edit/restore test; saved Take metadata stayed unchanged. Sampling, Run Storage contracts, V3.8X2/V3.9 workflow files and historical Releases are unchanged. GPU continuation was not tested for this repair. [Details and verification limits](REVIEW_RESTORATION_REPAIR.md). + +## Archived implementation notes (pre-V3.8; not the current guide) + +The version-labelled sections below are retained temporarily as source history only. They do not describe the V3.8 standard workflow, public surface, or recommended wiring. For an older workflow, use the migration policy and its matching historical Release/tag. + +### V3.7 High-Resolution Refinement Foundation + +V3.7 adds **H3 Continuum Sampler V3.7** and completes two foundations for resolution-changing Second Pass workflows without changing V3.6 Production defaults or the existing SIGMAS socket. + +### Conditioning survives a resolution change + +The Conditioning Adapter rebuilds First/Last images, continuation context, and an owned Still Image Guide at the target latent geometry. Physical-group ownership, Long Terminal Merge, Core PackedLayout/RoPE reconstruction, and first-pass Audio passthrough are preserved. Correctness gates passed at 704×704 and 1152×1152; larger canvases still require substantially more memory and processing time. + +### RefineSchedule gives refinement ranges explicit meaning + +The internal, versioned `RefineSchedule` contract supports `External`, `Full`, `Tail`, and `Partial`. `External` preserves the incoming SIGMAS tensor exactly, while `Tail` and `Partial` select exact ranges from the same source schedule. The accepted internal Tail 6 gate reproduced the six-evaluation suffix and preserved decoded Audio PCM bit-exact. This contract is an implementation foundation rather than a new public widget, and Production defaults remain unchanged. + +### Tail 6 cuts Second Pass time by about one third + +The largest practical benefit in V3.7 is a shorter high-resolution Second Pass when the internal schedule uses Tail 6. Instead of evaluating all ten points, Tail 6 runs the exact low-sigma suffix that performs the final refinement work. + +| Second Pass range | Sampling time | +|---|---:| +| Tail 10 | **100.27 s** | +| Tail 6 | **64.18 s** | +| Saved | **36.09 s / 36.0%** | + +In this gate, Tail 6 finished in about 64% of the Tail 10 time: roughly 1 minute 4 seconds instead of 1 minute 40 seconds. A separate Learned Latent Upscaler gate measured `121.17 s -> 66.32 s` for Tail 10 versus Tail 6, about 45% lower, while API total changed from `170.70 s` to `119.61 s`, about 30% lower. These measurements are configuration-specific and are not universal speed guarantees. + +`RefineSchedule` does not make each model evaluation faster. It gives Continuum an exact, versioned way to manage and execute the Tail 6 range; the speedup comes from running six evaluations instead of ten. Tail 6 is not enabled automatically, and V3.7 does not change Production defaults. + +### Still Image Guide remains Experimental + +V3.7 can map one Still Image Guide from an absolute frame to its owning physical group and rebuild it from the original image during a high-resolution Second Pass. Its positioning, Terminal Merge ownership, Run Storage identity, and non-owner isolation passed correctness testing. Core Add Guide uses hard-anchor semantics, however, so a Guide can cause an abrupt trajectory change at the anchor and can redirect later motion. **Still Image Guide is Experimental and remains on Production HOLD.** It should not be treated as a smooth transition control. + +### V3.6.1 Maintenance Hotfix + +V3.6.1 fixes focused fail-open/compatibility issues without changing the accepted Balanced 22 Masked AV path. Mixed Timeline/List prompts now report `H3C-P105`; Timeline parsing ignores only standalone `---` lines, keeps existing uncovered-chunk fallback, and never stops generation because of prompt syntax. + +With `Continuation Backend = Standard` and Audio Continuity enabled, `Balanced — 22 frames` continues to use Masked AV. `Fast — 5 frames`, `Strong — 39 frames`, and `Auto — conservative` now resolve to the accepted Reference Context transport before Run Storage identity is created. The selected UI values are not rewritten, and the status report states the resolved transport and reason. Standard with Audio Continuity disabled remains Masked Video; Compatibility remains Reference Context. + +Run Storage now treats every absent Last Frame spelling as the same empty identity. A normal no-Last-Frame run can therefore extend from two to three chunks as `2 reused, 1 generated`, including narrowly compatible V3.6.1 caches whose former final chunk stored `none`. A real connected Last Frame and the Long Terminal Merge atomic pair remain strict. Turning Run Storage Off also resets the hidden Regenerate From and Variation Nonce values to `Auto` and `0` before queueing. + +### V3.6 Masked AV Continuation + +![H3 Continuum V3.6 Masked AV Continuation](images/v36-masked-av-continuation.png) + +V3.6 adds **H3 Continuum Sampler V3.6**. Its Standard continuation backend places the previous finalized Video and Audio latent prefixes directly inside the next H3 target and protects them with Core noise masks. Unlike the V3.5 Reference Context route, it does not append the same prior frames as a separate Reference block, reducing the packed sequence processed by H3. + +| Continuation Backend | Purpose | +|---|---| +| `Standard` | Recommended V3.6 path. Balanced 22 + Audio Continuity uses Masked AV; Fast 5, Strong 39, and Auto safely fall back to Reference Context. When Audio Continuity is off, Masked Video is used. | +| `Compatibility` | Advanced fallback using the accepted V3.5 Reference Context behavior. | + +Internal transport identifiers are not exposed in the UI. Run Storage identity follows the resolved transport: Balanced 22 Masked AV stays separate from Reference Context, while a non-Balanced Standard fallback safely shares the identical accepted Reference contract. `Regenerate From` treats the Long Terminal Merge logical pair `[2,3]` as one atomic physical sample. + +### Accepted behavior + +- T2VA, I2VA, Reference Image + Reference Audio, Balanced 22-frame Joint AV, and 3×5-second FL2VA Long Terminal Merge GPU gates passed. +- Protected Video and Audio prefixes finalize bit-exact; generated regions remain owned by the sampler. +- The accepted Reference Image + Reference Audio output passed subjective audio listening with no reported click, dropout, or stereo-positioning defect. +- In the tested FL2VA Terminal Group 2 pairs, Standard removed 2,874 packed rows (`8.01%`) and reduced median Sampling time by `4.06%` versus Compatibility. Performance varies by model, hardware, and workflow. +- `chunk_seconds` now accepts 4.0–30.0 seconds with a 5.0-second default and 0.1-second step. The 5–15 second range remains recommended and validated; longer high-resolution chunks can substantially increase VRAM use and runtime. + +In the historical V3.6 package, V3.5.3, V3.5, and V3.4 Node IDs remained registered for saved workflows. V3.5.3 retained its original Reference Context behavior rather than being silently redirected to the V3.6 backend. This is not a V3.8 registration guarantee. + +## V3.5.3 Maintenance Hotfix + +V3.5.3 is a distribution-integrity maintenance release. It repairs the public Conditioning Bridge workflow's Width/Height links, removes stale orphan link IDs from the published V3.4/V3.5 templates, corrects Hybrid `` warning numbering, and repairs one legacy PNG for strict image decoders. + +There are no changes to generation behavior, nodes, sockets, Sampling, Conditioning payloads, Terminal Merge, Assembly, Seam, Run Storage, Prompt/CLIP caching, Video Guide optimization, or V3.4 compatibility. Existing V3.5.x workflows continue to load unchanged. + +## V3.5.2 Stabilization & Optimization Update + +V3.5.2 adds no new generation mode. It stabilizes the V3.5.x release, removes repeated Prompt/CLIP work when safe, and lowers temporary memory for long Video Guide inputs while preserving existing workflow and generation contracts. + +![H3 Continuum V3.5.2 stabilization and optimization results](images/v352-stabilization-optimization.png) + +The figure records the accepted optimization measurements. The final packaging gate additionally passed on ComfyUI 0.33.3 with a 172-entry Manifest, including this image. + +### What changed + +1. **Repeated-run Prompt/CLIP cache** — unchanged T2VA, I2VA, and FL2VA conditioning can reuse CPU-cached Prompt/CLIP results. Reference, scheduled, tokenizer-option, and hook-modified CLIP paths conservatively bypass the cache. Sampling still runs normally. +2. **Video Guide memory optimization** — the complete source remains part of finite validation and SHA-256 identity, but only the required prefix remains stored for VAE/conditioning work. +3. **Stabilization audit** — measured Sampling, Decode, Assembly, saving, optional-feature tax, and retained memory. No speculative Sampling, Driving Audio, Audio hash, Assembly, Seam, Session, or V3.4 optimization was adopted. + +### Measured results + +| Area | Before | V3.5.2 | Result | +|---|---:|---:|---:| +| T2VA repeated Prompt/CLIP | 5.898398 s | 0.000028 s | Cache HIT, `encode_calls=0` | +| FL2VA repeated Prompt/CLIP | 21.642687 s | 0.007128 s | Cache HIT, `encode_calls=0` | +| Video Guide peak additional RSS | 396.8 MiB | 114.7 MiB | About 71% lower | +| Video Guide retained storage | 225 MiB | 93 MiB | About 59% lower | + +Prompt/CLIP figures measure only the conditioning subphase, not total generation time. The Video Guide memory benchmark uses deterministic 300×256×256 RGB input with a 124-frame required prefix. A separate GPU A/B used an actual 15-second / 360-frame Video Guide and produced identical decoded video and audio PCM. + +The measured Sage-only production baselines on the tested RTX 5060 Ti 16 GB / 64 GB system were 168.069 seconds for 1×5-second 576×576 T2VA and 379.765 seconds for 3×5-second 640×640 FL2VA Long Terminal Merge. These are configuration-specific baselines, not universal speed guarantees. Sampling remained the dominant cost; Continuum Assemble + Seam stayed below 1%. + +> **V3.8.0 is the historical release baseline.** V3.8X2 remains a separate package-3.8.3 workflow, while V3.9 source is available on `main`. Historical implementation modules remain in source because V3.8X2 reuses them internally. The current registration has ten V3.8X2 IDs plus two V3.9 IDs. Use the matching historical Release/tag for workflows requiring other IDs. Still Image Guide remains Experimental. + +## V3.5.1 Reference Audio & Compatibility Update + +V3.5.1 added two focused features without changing the V3.4 Sampling, Conditioning, Terminal Merge, Assembly, Seam, or Run Storage contracts: + +1. **Optional Reference Audio** — native H3 audio conditioning that does not replace the generated final audio. +2. **Conditioning Bridge V3.5** — one complete Core-compatible `MODEL` and `CONDITIONING` object per Continuum physical group for external sampler workflows. + +The Reference Audio sockets are permanently defined by Python `INPUT_TYPES`. Dynamic socket changes and the UI-only `Hidden / Show` control were removed to prevent workflow save/reload value shifts. Node IDs, backend keys, Sampling, Conditioning, standard Seed handling, and other widgets are unchanged. + +In the historical V3.5.1 package, V3.4 node IDs and backend socket keys remained registered for saved-workflow compatibility. V3.4/V3.5 workflows continued to load there; V3.5.1 clarified the displayed input names without rewriting saved links. V3.8 instead uses the seven-node support boundary above. + +### Reference inputs at a glance + +The displayed names intentionally describe different jobs. They are not interchangeable pairs: + +| Displayed input | Connect from | Purpose | Final audio behavior | +|---|---|---|---| +| `reference_image_1`–`reference_image_3` | Still-image `IMAGE` | Persistent identity, subject, or appearance references | No audio | +| `Video Guide Frames` | Video loader `IMAGE` frame batch | Persistent motion, framing, timing, and appearance guidance across chunks | Does not carry the video's audio | +| `Driving Audio` + `Driving Audio VAE` | Audio loader, or video loader `AUDIO`, plus the matching audio VAE | Audio guidance whose effective source stream is preserved for final output | Replaces generated final audio with the selected source stream | +| `Reference Audio (Optional)` + `Reference Audio VAE (Optional)` | Standalone audio plus the matching audio VAE | Native H3 conditioning only | Does not copy or replace the generated final audio | + +For the common video-with-sound case, connect the loader's `IMAGE` output to `Video Guide Frames` and its `AUDIO` output to `Driving Audio`. Do not connect that audio to `Reference Audio (Optional)` unless conditioning-only behavior is specifically intended. + +![Sampler V3.5 permanent Reference Audio, Video Guide Frames, Driving Audio, and Reference Image inputs](images/v351-video-guide-frames.png) + +`Reference Audio (Optional)` and `Reference Audio VAE (Optional)` are always present exactly as defined by the node's Python input schema. V3.5.1 no longer adds or removes these sockets dynamically and has no frontend-only `Hidden`/`Show` widget. This keeps positional widget values aligned when workflows are saved and reloaded; backend input keys and existing workflow links are unchanged. + +![Sampler V3.5.1 with permanent Reference Audio sockets and aligned widgets](images/v351-reference-audio-permanent.png) + +### Conditioning Bridge V3.5 + +`H3 Continuum Conditioning Bridge V3.5` is the Advanced connection point for external sampling. Connect `model`, `clip`, the externally processed `video_latents`, `assembly_plan`, and `refine_context`; connect `video_vae` only when the selected conditioning path requires it. The node returns parallel `group_models` and `conditioning` lists whose length equals the physical-group count. Each list item remains a complete ComfyUI object—conditioning entries are never flattened into the physical-group list. + +```text +H3 Continuum Sampler V3.5.video_latents + -> external H3 latent processor (for example LBH) + -> H3 Continuum Conditioning Bridge V3.5 + -> Core BasicGuider + external sampler + -> Core Video / Audio VAE Decode + -> H3 Continuum Assemble + Seam V3.5 +``` + +The external workflow owns AV LATENT pairing, noise, SIGMAS, Audio Lock, sampling, and audio passthrough. The Bridge only exposes the prepared `MODEL` and `CONDITIONING` objects plus the updated Assembly Plan. + +![V3.5.1 LBH and Conditioning Bridge external sampling flow](images/v351-lbh-conditioning-bridge-flow.svg) + +The retired V3.5.1 LBH + Conditioning Bridge example is available only from its matching historical Release/tag; it is not a V3.8 workflow. + +The example keeps the standard Core titles `BasicGuider`, `BasicScheduler`, and `SamplerCustomAdvanced`. Public examples do not rename Core nodes, so they remain immediately distinguishable from Continuum and third-party nodes. The workflow also uses external nodes for LBH latent upscaling, AV LATENT concatenation/separation, loading/saving, and optional acceleration; install or replace those nodes according to your ComfyUI environment. + +LBH changes latent geometry only. It has no Continuum denoise-strength control: use the external `BasicScheduler` SIGMAS to decide how strongly the resized latent is regenerated. A larger LBH scale such as 1.5x may still be fast, but memory, decode, and sampling costs increase with the target canvas. + +## What's new in V3.5 + +V3.5 has two major additions: + +1. **Continuum-aware Second Pass / Hi-Res Fix** — refine externally processed H3 latents or use the integrated pixel/VAE 2x path without changing V3.4 sampling. +2. **Low-memory Assemble + Seam** — write the final video IMAGE directly to RAM or a Windows-safe mapped file while preserving Exact Duration, Seam, Terminal Merge order, and audio behavior. + +### 1. Second Pass and Hi-Res Fix + +**H3 Continuum Hi-Res Fix V3.5** is the one-node Main path. It performs Video VAE Decode, bounded pixel resize, Video VAE Encode, and one low-denoise Second Pass. The integrated Main path remains **Experimental** because long 2x runs can exceed GPU memory. + +![H3 Continuum Hi-Res Fix V3.5 node](images/v35-hires-fix-node.png) + +Normal Hi-Res Fix wiring: + +```text +H3 Continuum Sampler V3.5 + -> H3 Continuum Hi-Res Fix V3.5 + -> Core Video / Audio VAE Decode + -> H3 Continuum Assemble + Seam V3.5 +``` + +**H3 Continuum Second Pass V3.5** is the stable Advanced bridge requested in Issue #8. It accepts an externally processed H3 video latent and samples each Continuum physical group once while preserving prompt policy, seed reporting, group order, and the original first-pass audio LATENT object. + +![H3 Continuum Second Pass V3.5 node](images/v35-second-pass-node.png) + +External latent processor wiring, such as LBH: + +```text +H3 Continuum Sampler V3.5.video_latents + -> external H3 latent processor + -> H3 Continuum Second Pass V3.5 + -> Core Video / Audio VAE Decode + -> H3 Continuum Assemble + Seam V3.5 +``` + +Connect the original `audio_latents`, `assembly_plan`, and `refine_context` from Sampler V3.5 to the Hi-Res Fix or Second Pass node. Use the same final MODEL and CLIP, plus a separate sampler and workflow-side SIGMAS for the refinement pass. The nodes do not contain an internal denoise value. + +The `BasicScheduler` is deliberately omitted from the screenshots. It is still required to create the Second Pass `SIGMAS`. The GPU-tested baseline is: + +```text +sampler: res_multistep +scheduler: simple +steps: 10 +denoise: 0.35 +``` + +Do not chain Hi-Res Fix and another Second Pass in the standard path: Hi-Res Fix already performs one Second Pass internally. `H3 Continuum Latent Resize V3.5` is an advanced utility, not the recommended Main 2x path; direct high-ratio interpolation of the 24-channel H3 latent produced persistent artifacts in testing. + +If Hi-Res Fix is not connected, V3.4 sampling remains on its original path. V3.5 does not add a forced CPU-offload or low-VRAM sampler mode. `Hi-Res Fix enabled=false` is a lazy exact passthrough and does not evaluate its resize, VAE, or Second Pass dependencies. + +See [V3.5 Hi-Res Fix and Second Pass workflow guide](V35_HIRES_FIX.md) for exact sockets, bypass behavior, external-upscaler wiring, and limits. + +### 2. Low-memory Assemble + Seam + +**H3 Continuum Assemble + Seam V3.5** adds manual RAM / Disk-backed selection and a conservative Auto policy. Disk-backed mode writes the final video IMAGE directly to a mapped file instead of rebuilding the entire completed video in anonymous RAM. Audio remains in RAM. + +| Backend | Behavior | +|---|---| +| Auto | Uses RAM only when the final IMAGE is at most 4 GiB and physical-memory reserve is sufficient; otherwise selects Disk-backed. Missing memory counters fail safely to Disk-backed. | +| RAM | Existing in-memory output behavior. | +| Disk-backed | Maps the final video IMAGE to a temporary file; video only. Audio remains in RAM. | + +Disk-backed mode preserves the backing file while ComfyUI still references the returned Tensor. Downstream nodes can still allocate a full RAM copy, so Continuum's low-memory guarantee applies to its own assembly stage, not arbitrary downstream implementations. + +#### Measured memory reduction + +The validated stress case assembled a deterministic `1536 x 1536`, 360-frame float32 IMAGE with a final size of `9.49 GiB`. RAM and Disk-backed outputs were video/audio hash-identical. + +| Windows process metric | RAM backend | Disk-backed | Reduction | +|---|---:|---:|---:| +| Private memory | 12.41 GiB | 2.91 GiB | **9.50 GiB** | +| USS | 10.29 GiB | 0.80 GiB | **9.49 GiB** | + +This is a **system-RAM/private-commitment** reduction during Continuum Assembly, not a claim that sampler GPU VRAM is reduced. Windows RSS includes resident mapped-file pages and was therefore similar between the two backends; private memory and USS are the relevant measurements. Disk-backed I/O can make final assembly slower, but it does not slow the preceding model sampling path. + +#### Optional Windows tip: disable pinned memory + +For MiniMax H3 on Windows, starting ComfyUI with `--disable-pinned-memory` can substantially reduce host-RAM pressure. In one local RTX 5060 Ti 16 GB / RAM 64 GB check, a `0.4 MP (640 x 640), 2 x 5s` run completed in `5m 23s`, with about `6.1 GB` sampler RSS and `8.6 GB` after assembly, while showing little practical speed difference from normal startup. + +```text +--disable-pinned-memory +``` + +This is an environment-specific tip, not a default requirement or a universal crash fix. Compare the same workflow and seed with only this option changed, and keep it only when memory use improves without a meaningful speed or stability regression. It disables ComfyUI's pinned host-memory path; it does not guarantee that Windows Shared GPU Memory will never be used. + +### GPU acceptance at a glance + +| Path | Tested condition | Result | +|---|---|---| +| Main Hi-Res Fix | FL2VA 1 x 5s, 576 -> 1152 | PASS | +| Main Hi-Res Fix | Hybrid FL2VA + Reference 1 x 5s, 576 -> 1152 | PASS | +| Advanced Second Pass | Hybrid/Reference 1 x 5s, 576 x 576 | PASS | +| Main Hi-Res Fix | FL2VA 3 x 5s, 576 -> 1152, RTX 5060 Ti 16 GiB | **CUDA OOM at terminal 77T group** | + +For the failed 3 x 5s case, First Pass and the 37T Second Pass group completed. The final logical chunks `[2,3]` remained one 77T Long Terminal Merge physical group and exceeded the tested 16 GiB GPU at its first 1152x1152 Second Pass inference. This is a documented resource limit; the Continuum grouping contract did not change. + +## V3.4 compatibility baseline + +The historical V3.5 package retained V3.4 nodes for saved workflows; it did not replace their Node IDs, public sockets, Sampling, Conditioning, Terminal Merge, Assembly, Seam, or Run Storage behavior. Those historical workflows remain available through the corresponding Release/tag. Current V3.9 exports only its twelve registered IDs, even when older implementation modules remain in source. + +![H3 Continuum V3.4 workflow overview](images/v34-workflow-overview.png) + +V3.4 focuses on the two reference workflows that are most useful in normal production: + +- **Driving Audio**: use an existing audio source as the preserved final audio while it guides the H3 generation across chunks. +- **Video Guide Frames**: use a video loader's IMAGE frame batch as persistent guidance for appearance, motion, framing, and timing without treating it as a frame-by-frame copy. +- **Hybrid FLF + Reference**: use First Frame, Last Frame, and persistent Reference Images together without adding a separate mode or model allowlist. +- **FL2VA Terminal Merge**: keep the final two 5-second FL2VA chunks as one physical 10-second sampling and decode unit for Core-equivalent Last Frame handling. +- **Restartable chunks**: reuse completed chunks with Run Storage and regenerate only the selected part when the generation contract remains compatible. +- **Core compatibility**: remove Continuum-only rejection of unknown upstream/custom nodes and remove the obsolete `strict_compatibility` control. +- **Spectrum interoperability**: use the official H3 Continuum Interop API v1 when Spectrum is installed; Spectrum remains optional. +- **Simpler public interface**: Timeline Video and the earlier timeline-audio paths are hidden from the V3.4 public sampler interface rather than being presented as stable production features. + +### Direction change: from timeline generation to preserved references + +Earlier development explored Timeline Video and timeline-audio conditioning. Those paths can be useful for experiments, but they are not the default production behavior: chunk boundaries can change visual content, and generated audio can diverge from a supplied song, dialogue, or effects track. + +V3.4 therefore prioritizes: + +1. **Driving Audio** for cases where the supplied audio should remain the final audio stream. +2. **Video Guide Frames** for cases where a supplied video should guide the generated result without requiring exact frame reproduction. +3. **Chunked generation and Run Storage** for practical retries and long-form work. + +This is a usability and reliability decision, not a claim that the experimental timeline paths are impossible. They are hidden from the stable interface while the public workflow stays focused on predictable user-controlled inputs. + +## Example workflow + +- [V3.9 workflow JSON](../examples/workflows/MiniMax_H3_Continuum_V39.json) and [ZIP](../examples/workflows/MiniMax_H3_Continuum_V39.zip) — current V3.9 graph; the ZIP contains only this JSON +- [V3.8X2 workflow JSON](../examples/workflows/MiniMax_H3_Continuum_V38X2.json) — retained V3.8X2 graph; Spectrum is disabled by default +- [V3.8X2 workflow ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2.zip) — contains both the V3.8X2 and V3.9 JSON files; it is not a custom-node installer +- [V3.8X2 Helper-explicit JSON](../examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.json) and [ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.zip) — the same graph under an explicit distribution name + +The declared Registry payload includes both V3.9 and V3.8X2 graphs; it is **not dependency-free**. Spectrum, rgthree, and KJNodes must be installed separately; ComfyUI-Easy-Use is not required. Historical workflow files remain in GitHub source but are excluded from the Registry payload. For an older saved workflow, use its matching historical Release/tag as described in the [migration policy](V38_RELEASE_AND_MIGRATION.md). Registry packaging validation/publication is separate from this GitHub source update. + +## V3.4 feature details + +### Driving Audio + +Driving Audio is the primary V3.4 audio path. + +- Source audio guides each chunk at its absolute sequence position. +- The original effective stream is preserved for final output. +- Generated audio and Audio Seam processing are bypassed while Driving Audio is active. +- Audio is not independently rewritten at every chunk boundary. +- Perfect lip sync is not guaranteed; results still depend on H3, source material, prompts, and sampling. + +### Video Guide Frames + +`Video Guide Frames` provides a persistent native H3 video guide across all chunks. Despite its ComfyUI `IMAGE` socket color, it expects the frame batch produced by a video loader rather than one ordinary still-image reference. + +- It guides motion, framing, timing, and appearance. +- It is not a direct pixel-copy or deterministic identity-transfer system. +- Video and audio are independent. Route a loader's IMAGE output to `Video Guide Frames` and AUDIO output to `Driving Audio`. +- Source decoding and frame-rate conversion remain loader responsibilities. Continuum does not impose an unnecessary forced 24 fps conversion. + +| Video Guide Size | Purpose | +|---|---| +| Efficient - 0.4 MP | Default; lower token cost and faster iteration. | +| Balanced - 0.6 MP | More detail at a higher compute cost. | +| Match Output | Largest reference input; potentially much slower and heavier. | + +#### Conditional size controls and internal resize + +The compact V3.5 UI shows each size control only when its corresponding input is connected: + +- Connect any of `reference_image_1` through `reference_image_3` to reveal `Reference Size`. +- Connect `Video Guide Frames` to reveal `Video Guide Size`. Earlier workflows and reports may call this control `Video Reference Size`; V3.5.1 uses the clearer display name `Video Guide Size` while preserving the existing backend key and saved-workflow compatibility. + +These controls are therefore not missing from a newly created node; they are hidden until relevant. Reference Images and Video Guide Frames are resized internally before VAE encoding. Video Guide Frames preserve their aspect ratio, align to the H3 canvas, and use Lanczos when reduction is needed. Smaller sources are not enlarged automatically. An external resize node is normally unnecessary, but remains useful for deliberate cropping, forced upscaling of a small source, or a custom resize/upscale algorithm. Video decoding and frame-rate conversion remain the loader's responsibility. + +### Hybrid First/Last Frame + Reference Images + +You need an hybrid version of MiniMax-H3 to make both I2V and reference images work: + +https://huggingface.co/smhfacct/Minimax-H3-fl2va-ref2va-hybrid-models + +https://www.reddit.com/r/StableDiffusion/comments/1vm62pj/i_tested_out_the_b2049_hybrid_variant_for_r2v/ + +V3.4 supports First Frame, Last Frame, and persistent Reference Images in the same Continuum run. The public sockets and controls are unchanged: connect the inputs that the selected H3 model supports. + +- Pure FL2VA and pure Ref2VA identity contracts remain unchanged. +- The combined path receives a distinct hybrid identity so Run Storage does not reuse incompatible chunks. +- Prompt/reference mismatches produce guidance warnings rather than a Continuum-only execution stop. +- No model-name allowlist or automatic model replacement is used. + +Local checks passed with the B2049 hybrid variant. FL-only models could run, but showed less reliable transitions when FLF and Reference Images were combined; a hybrid-capable model is therefore recommended for this path. + +### FL2VA Terminal Merge + +For 5-second FL2VA workflows, Continuum keeps the final two logical chunks together as one physical sampling and VAE decode unit. + +- `2 x 5s`: one 243-frame physical sample, decoded once, then adjusted to the requested 240-frame output. +- `3 x 5s` and longer: earlier chunks use normal Continuation; the final two chunks use one 260-frame physical sample containing the 22-frame continuation context. +- After the terminal physical decode, the 22-frame context prefix is removed. For `3 x 5s`, this gives `124 + 238 = 362` frames before exact-duration adjustment to 360 frames. +- If the final two prompt sections differ, they are rebased internally to a local `[0-5s]` / `[5-10s]` timeline. Identical prompts retain the accepted shared-prompt path. +- The terminal pair uses one physical seed and one physical decode while Run Storage continues to preserve normal logical chunk entries. + +Local GPU comparison confirmed that `2 x 5s` matches Core 10-second FL2VA sampling and terminal decode behavior. A `3 x 5s` Timeline run also passed with three logical chunks, two physical sampling passes, two physical decode groups, and a 15.000-second output. The short still period near the supplied Last Frame was also present in the equivalent Core result and is treated as normal FL2VA convergence rather than a Continuum-specific failure. + +### Core-first, permissive execution + +V3.4 removes or relaxes Continuum-only rejection rules that blocked otherwise runnable Core H3 experiments. + +- Prompt problems prefer warnings and fallback behavior instead of stopping. +- Unknown upstream/custom wrapper classes are not blanket-rejected. +- There is no model allowlist and no automatic model replacement. +- The obsolete strict_compatibility control is removed from the V3.4 interface. +- Current Core PackedLayout behavior is supported without requiring the removed legacy frame_count argument. + +Hard stops remain only for states that cannot execute safely, including corrupt latent topology, incompatible assembly data, invalid persisted revisions, or unusable mandatory payloads. + +### Cleaner interface + +The public nodes focus on normal production controls. Developer diagnostics and detailed reports are available through ComfyUI settings. The current release path uses `H3 Continuum Sampler V3.8`; older sampler and assembler notes below are retained for historical compatibility context. + +H3 Continuum Sampler V3.7 with optional Guide input + +![H3 Continuum Assemble + Seam V3.5](images/v35-assemble-seam-node.png) + +### Timeline paths + +Experimental Timeline Video and earlier timeline-audio paths remain in historical implementation code but are not exported by V3.8. Open V3.3 workflows with their historical package; new V3.8 workflows should use `Driving Audio` and `Video Guide Frames`. + +## Installation + +V3.8 is verified against ComfyUI 0.34.2. Historical acceptance records below may mention older ComfyUI versions; they are not the current installation target. + +### Updating an existing installation + +For a Git checkout, run the update from the custom-node directory: + +```powershell +cd ComfyUI/custom_nodes/ComfyUI-H3-Continuum +git pull --ff-only origin main +``` + +Restart ComfyUI after the update. If the node was installed with ComfyUI Manager, use its **Update** action instead of running `git pull` manually. Do not mix Manager updates and a separate Git checkout for the same installation. + +After the backend restart, search for `H3 Continuum Sampler V3.8` with the V3.8X2 workflow, or `H3 Continuum Sampler V3.9` and `H3 Continuum Reference Images V3.9` with the V3.9 workflow. The source registers twelve IDs in total. If they are missing, check the startup console for the `H3 Continuum 3.9.0 loaded` message and any `ComfyUI-H3-Continuum` import error. + +Search for H3 Continuum or Continuum in ComfyUI Manager, or install manually: + +~~~bash +cd ComfyUI/custom_nodes +git clone https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum.git +~~~ + +Restart ComfyUI after installation or update. + +## V3.5 nodes + +### H3 Continuum Sampler V3.5 + +The V3.5 sampler preserves V3.4 generation behavior and adds `refine_context` as the sixth output. Connect it to V3.5 Hi-Res Fix, Second Pass, or Conditioning Bridge paths that need the original conditioning context. V3.5.1 also provides permanently visible optional Reference Audio sockets. + +### H3 Continuum Conditioning Bridge V3.5 + +Advanced external-sampler bridge. It returns aligned `group_models` and complete `conditioning` objects per physical group together with the updated Assembly Plan. It does not create AV LATENT pairs, noise, SIGMAS, Audio Lock, or external sampling policy. + +### H3 Continuum Hi-Res Fix V3.5 (Experimental) + +Main convenience node. When enabled, it performs the Video VAE pixel-resize round trip and Second Pass internally. When disabled, video/audio LATENT objects and the Assembly Plan pass through unchanged and the sampling/VAE dependencies remain lazy. + +### H3 Continuum Second Pass V3.5 + +Advanced Issue #8 bridge for an externally processed video latent. It samples once per physical group, reads the physical prompt contract from the Assembly Plan, and returns the original first-pass audio LATENT objects. + +### H3 Continuum Latent Resize V3.5 + +Utility that changes only video-latent H/W. It does not sample. Direct 2x interpolation at large square targets is not the accepted Main Hi-Res Fix method. + +### H3 Continuum Assemble + Seam V3.5 + +V3.4-compatible assembly with Auto, RAM, and Disk-backed video-buffer backends. Exact Duration and seam corrections write directly to the selected target buffer. + +## V3.4 nodes + +### H3 Continuum Sampler V3.4 + +Main inputs: + +- model, clip, video_vae, sampler, sigmas +- Sequence Prompt +- first_frame, last_frame +- reference_image_1 through reference_image_3 +- Video Guide Frames +- driving_audio, audio_vae + +Outputs: + +- video_latents, audio_latents +- assembly_plan, status +- driving_audio + +Visible controls: + +- Prompt Format, chunks, chunk_seconds +- width, height, continuity +- base_seed, control after generate +- audio_continuity +- Run Storage +- Reference Size +- Video Guide Size + +`chunk_seconds` uses one shared duration for every chunk. The default is 5.0 +seconds; 5–15 seconds remains the recommended and validated range. Durations up +to 30.0 seconds are supported, but values above 15 seconds can substantially +increase VRAM use and processing time, especially at high resolution. Per-chunk +variable durations are not supported. + +### H3 Continuum Assemble + Seam V3.4 + +Inputs: images, audio, assembly_plan, and driving_audio. + +Controls: Audio Seam and Video Seam. + +When Driving Audio is connected, preserved source audio is selected for final output and generated audio seam processing is bypassed. + +## Connection order + +~~~text +H3 model / CLIP / Video VAE / sampler / sigmas + | + v + H3 Continuum Sampler V3.4 + | | | + video_latents | assembly_plan + | + audio_latents + +video_latents -> Core VAE Decode ------- images --+ +audio_latents -> Core VAE Decode Audio -- audio --+--> H3 Continuum Assemble + Seam V3.4 +assembly_plan -------------------------------------+ +driving_audio -------------------------------------+ +~~~ + +For a source video with sound: + +~~~text +Video loader IMAGE -> Video Guide Frames +Video loader AUDIO -> driving_audio +~~~ + +The supplied templates use Video Helper Suite for this split. Any compatible IMAGE/AUDIO loader may be substituted. + +## Prompt formats + +### Timeline + +For **Timeline** mode, each time-range header must be written on its own line. Write the scene description on the following line or lines. + +Recommended format: + +~~~text +[0-5s] +First section: action, camera movement, and scene progression. + +[5-10s] +Continue from the exact final state of the previous section. + +[10-15s] +Continue naturally without resetting the scene. +~~~ + +For 10-second chunks: + +~~~text +[0-10s] +Describe the first scene. + +[10-20s] +Describe the next scene. + +[20-30s] +Continue the sequence. + +[30-40s] +Describe the final section. +~~~ + +**Important:** do not place the prompt text on the same line as the time header. + +Correct: + +~~~text +[0-5s] +Describe the scene. +~~~ + +Not recommended: + +~~~text +[0-5s] Describe the scene. +~~~ + +The second form is not recognized as a Timeline header. When **Prompt Format = Auto**, it may therefore be interpreted as a Fixed prompt instead, causing the complete text to be reused across chunks. + +The time ranges should normally match the configured `chunk_seconds`. + +- `chunk_seconds = 5` → `[0-5s]`, `[5-10s]`, `[10-15s]` ... +- `chunk_seconds = 10` → `[0-10s]`, `[10-20s]`, `[20-30s]` ... +- `chunk_seconds = 15` → `[0-15s]`, `[15-30s]`, `[30-45s]` ... +- `chunk_seconds = 30` → `[0-30s]`, `[30-60s]`, `[60-90s]` ... + +#### Global preamble and per-chunk prompting + +Text placed before the first Timeline header is treated as a **global preamble** and is automatically included in every chunk prompt. This is a good place for information that should apply throughout the sequence, such as subject definitions, overall style, persistent environment details, or other global instructions. + +However, the Timeline sections do not need to contain only the changing action or scene description. For complex scenes, multiple characters, or longer chunks such as 10–15 seconds, it can be useful to repeat important details about the subjects, appearance, location, camera, and current scene state inside each section. More self-contained chunk prompts may sometimes give H3 stronger continuity and reduce cut-like transitions or drift. + +Continuum carries visual/audio context from the previous chunk, but that does not guarantee that H3 will preserve every semantic detail automatically. Instead of relying only on phrases such as "continue from the previous chunk," describe the intended next action and important continuity details explicitly. + +For example: + +~~~text +Subject: A woman in a red jacket. +Style: cinematic nighttime photography. + +[0-15s] +A woman in a red jacket stands at the nightclub bar, speaking with two people beside her. Medium shot, slow handheld camera movement, crowded dance floor visible behind them. + +[15-30s] +The same woman in the same red jacket continues the conversation at the same bar with the same two people. The camera remains close and slowly moves around the group as one person replies. + +[30-45s] +The woman turns toward the dance floor while remaining beside the bar. Keep the same clothing, nightclub environment, nearby characters, and continuous camera style. +~~~ + +The global preamble is optional. If fully self-contained prompts inside every Timeline section give better continuity for a particular model or scene, that is also a valid approach. + +### Other prompt formats + +**List** prompts use `---` separators. + +**Fixed** reuses one prompt for every chunk. + +**Auto** detects Timeline, List, or Fixed formatting from the supplied text. + +If Timeline syntax cannot be parsed, V3.4 reports a warning and falls back to an applicable prompt mode rather than rejecting an otherwise runnable workflow. + +## Run Storage + +Run Storage preserves completed chunks and can reuse them after restarting ComfyUI. Regenerate From selects the first chunk to regenerate; earlier compatible chunks remain unchanged. + +![Regenerate From](images/v34-regenerate-from.png) + +### Keep the first part and regenerate the second part + +For a normal T2VA or I2VA run configured as `2 chunks x 5 seconds`, Chunk 1 is the first five seconds and Chunk 2 is the second five seconds. + +Before the first generation: + +1. Set `Run Storage` to `Save + Auto Resume`. +2. Choose a fixed `Run Name` and keep the Base Seed fixed. +3. Generate the complete ten-second video once. Continuum stores the raw Video/Audio latent for each completed chunk. + +To generate a different second part while keeping the first part: + +1. Keep the same Run Name, Base Seed, model, LoRA, resolution, sampler, SIGMAS/steps, references, and Continuation settings. +2. Set `Regenerate From` to `Chunk 2`. +3. Leave `Variation Nonce = 0` for automatic new variations after completed regenerations, or choose a different positive value for an explicit variation. A compatible interrupted regeneration at `0` resumes its existing variation; reusing the same positive value does not request a different result. +4. Queue the workflow again. + +Continuum loads the stored Chunk 1 instead of Sampling it again, uses its finalized end as the continuation context, and Samples a new Chunk 2. The report should show `1 reused, 1 generated`. The complete output is then decoded and assembled again; the saved first chunk is reused, while boundary Seam processing is evaluated during the new assembly. + +If the two parts need different instructions, define them with List or Timeline Prompt syntax before the first generation. Changing the Prompt Plan or other generation-contract inputs after a stored run can create a new revision, in which case the previous Chunk 1 may not be reusable. A run originally generated with `Run Storage = Off` cannot be reused retroactively. `Regenerate From` works at chunk boundaries; it does not edit an arbitrary region inside one chunk. + +FL2VA Long Terminal Merge is an important exception: its final two logical chunks are one atomic physical Sampling unit, so those two chunks are regenerated together. + +- Use a fixed seed for reproducible resume. +- Changing persistent models, LoRAs, references, prompts, resolution, or sampling settings can create a new revision. +- Unknown upstream node classes no longer cause rejection by themselves; reuse still depends on compatible observable contracts and hashes. + +## Spectrum interoperability + +Spectrum is optional. Current Spectrum releases officially support H3 Continuum Interop API v1. + +- Chunk 1 uses the normal initial path. +- Chunk 2 and later request Actual Prefix 2. +- Successful continuation logs accepted H3 Continuum API v1, actual prefix=2. +- Unsupported Spectrum versions fall back without a private patch. + +See [Spectrum v0.2.15 H3 Continuum interoperability](https://github.com/xmarre/ComfyUI-Spectrum-MiniMax-H3#v0215-h3-continuum-interoperability). + +Turbo LoRA and Spectrum are not mutually exclusive. Quality and speed remain workflow-dependent. + +## Issue and pull-request response + +### Issue #3 + +V3.4 removes blanket rejection of unknown upstream/custom class names. Run Storage evaluates the observable generation contract instead of treating an unfamiliar wrapper as automatically incompatible. + +See [Issue #3](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/3). + +### Issue #4 + +V3.4 follows the current Core H3 layout contract and no longer requires the legacy frame_count parameter. The old strict-compatibility toggle is not part of the public interface. + +See [Issue #4](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/4). + +### Issue #7 + +V3.4 allows First Frame, Last Frame, and Reference Images to be used together without requiring the separate BigStationW node or introducing a model allowlist. Hybrid presentation ordering and Run Storage identity are handled by Continuum while pure FL2VA and pure Ref2VA behavior remain unchanged. + +The combined path was exercised locally with the B2049 hybrid variant. FL-only models were less consistent for this specific combination, so the README recommends a hybrid-capable model rather than rejecting other models in code. + +See [Issue #7](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/7). + +### Pull request #1 and #2 + +These older partial pull requests were superseded by the consolidated [Pull request #5](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/pull/5). V3.4 selectively includes the compatibility direction needed for the current Core H3 contract, while the broader upstream synchronization and release-automation scope remains a separate upstream review. + +The older pull requests are not represented as merged V3.4 changes. + +### Pull request #5 + +PR #5 is the consolidated upstream-to-fork proposal covering current H3 compatibility, reference-input handling, assembly memory behavior, exact-duration handling, continuation-reference interoperability, and CI/release automation. V3.4 already contains the user-facing stability decisions described above; the PR remains an upstream review item and is not claimed as merged here. + +## V3.4 input connection patterns + +V3.4 separates the video-guide frame input from the driving-audio input. Choose the connection pattern that matches your source material. + +### 1. Audio only + +Connect `Load Audio` to `Driving Audio`. Use this when an existing song, dialogue track, or sound effect should remain the final audio. `Video Guide Frames` is not required. + +![Driving Audio connection](images/v34-driving-audio-connection.png) + +### 2. Video with its own audio + +Connect `Load Video (Upload)` `IMAGE` to `Video Guide Frames`. If the uploaded video contains the audio you want to preserve, connect its `AUDIO` output to `Driving Audio` as well. + +![Video Guide Frames and embedded Driving Audio connection](images/v34-video-reference-with-audio.png) + +### 3. Video and audio from separate sources + +Connect `Load Video (Upload)` `IMAGE` to `Video Guide Frames`, then connect a separate `Load Audio` node to `Driving Audio`. Use this when the video guide and the final audio source are different files. + +![Separate Video Guide Frames and Driving Audio connection](images/v34-video-reference-separate-audio.png) + +Both inputs are optional. Connect `Video Guide Frames` when video guidance is needed, and connect `Driving Audio` when the supplied audio should be preserved in the final output. + +### Video Guide Frames frame rate + +Use a 24 fps source for `Video Guide Frames`. `Load Video (Upload)` may accept files recorded at 25 fps or another frame rate, but acceptance alone does not guarantee correct temporal alignment with H3. For a non-24 fps source, set `force_rate` to `24` in `Load Video (Upload)`, or convert the file to 24 fps before loading it. If the source is already 24 fps, leave `force_rate` at its default and do not resample it. + +## Current validation status + +The public-surface suite checks ten V3.8X2 IDs and twelve IDs including V3.9, official workflows and matching ZIP payloads, declared external dependencies, Registry exclusions, preserved V3.8 widget/socket order, and presentation-only `Show Advanced Settings` / `Hide Advanced Settings` behavior. Registry payload hashes are recorded in `REGISTRY_MANIFEST.sha256`; source integrity is recorded in `MANIFEST.sha256`. + +**Historical V3.8X2 preparation evidence:** the full CPU suite passed **1,367 passed / 1 skipped / 0 failed**. The prior dedicated GPU Functional Gate passed a `3 × 5 second` Review Each Chunk run: Q1–Q3 generated one physical group at a time, and Q4 reused all three groups with `3 reused / 0 generated`. Q3 and Q4 had identical decoded RGB and PCM SHA-256 values. The separate nine-reference Gate is recorded below. These checks confirm functional execution, prefix reuse, AV reconstruction, and the tested nine-reference path; they are not a full subjective image- or audio-quality rating. + +### Nine-reference input gate (0.3 MP reference images) + +On the tested NVIDIA GeForce RTX 5060 Ti 16 GB / 64 GB RAM system, a `3 × 5 second` Full Run completed with Reference Images 1–9, `512 × 608` output, Spectrum Off, LoRA Off, Sage Attention, Balanced 22-frame continuity, and test reference images sized to about `0.30 MP` each. Sampling used approximately `15.0–15.5 GiB` of the available `16,311 MiB` VRAM, so this is a high-resource configuration with little headroom on a 16 GB GPU. + +Use the nine-reference bundle when the extra appearance or pose evidence is needed and the system has enough VRAM headroom. This single-seed gate confirms execution and basic visual continuity; it does not establish that nine references are universally higher quality than three, nor guarantee operation with higher-resolution references, larger output sizes, different models, or other concurrent GPU workloads. + +For nine-reference use on a 16 GB GPU, pre-resize copies of the source images to around 0.30 MP before loading them. Internal conditioning resize remains active, but it should not be treated as a substitute for controlling the memory pressure of nine large source images. + +Validation results apply to the tested local source and environment. They do not guarantee that an installed copy is current, that all models/wrappers fit every GPU, or that a different prompt will have the same visual quality. + +## Limits + +- V3.8 Second Pass / `refine_context` is not supported while a Review sequence is partial. V3.9 supports validated output-group conditioning for partial Review as described above. +- The 16GB GPU Gate passed with small VRAM headroom; memory use depends on the complete environment and workflow. +- Continuation does not guarantee frame-perfect identity or motion. +- Video Guide Frames guides H3; it does not reproduce every source frame. +- Driving Audio preserves selected audio, but visual lip synchronization remains model-dependent. +- FL2VA may settle on the supplied Last Frame before the requested duration ends; equivalent Core runs can show the same behavior. +- Long, high-resolution sequences may exceed system RAM during final decode and assembly even when chunk sampling succeeds. +- Disk-backed assembly lowers anonymous/private memory commitment for the final Continuum IMAGE, but Core Decode and downstream nodes may still allocate large RAM copies. +- Main Hi-Res Fix 3x5 2x is not accepted on the tested RTX 5060 Ti 16 GiB: the 37T group completed at 1152x1152, but the terminal 77T group failed at its first Second Pass inference with CUDA OOM. Reference/Hybrid-specific 1x5 acceptance passed; longer Hybrid/Reference cases remain unverified. +- Direct high-ratio interpolation with Latent Resize can create persistent H3 artifacts; use the Pixel/VAE Main path or an appropriate external H3 latent processor. +- Match Output can be substantially slower than 0.4 MP or 0.6 MP. +- Seam correction may keep the native boundary when a proposed correction is not safer. +- Optional template nodes must be installed, replaced, or bypassed by the user. + +## License + +See [LICENSE](../LICENSE). diff --git a/docs/README_HISTORY_JA.md b/docs/README_HISTORY_JA.md new file mode 100644 index 0000000..ae1a0b5 --- /dev/null +++ b/docs/README_HISTORY_JA.md @@ -0,0 +1,517 @@ +# H3 Continuum — 日本語の更新履歴・実装資料 + +[現在の日本語README](../README_JA.md) · [日本語の詳細ガイド](USER_GUIDE_JA.md) · [CHANGELOG](../CHANGELOG.md) + +正本`84899c1`(パッケージ3.9.8)の日本語READMEにあった更新説明・旧版画像・検証記録を保持しています。 +「現行」「main上」などの表現、古い導入コマンド、版ごとの受入結果は当時の文脈です。 +現在の導入案内や、未検証の条件を含む包括的な合格として読み替えないでください。 +最新版の入手先と対応範囲は、日本語READMEを参照してください。 + + + + +**最新版パッケージは3.9.8です。** 固定版のGitHub配布は下表を参照してください。Registryの公開状態は別に確認します。 + +| バージョン | Release | カスタムノード本体ZIP | SHA-256チェックサム | +| --- | --- | --- | --- | +| 3.9.8 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.8) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.8/ComfyUI-H3-Continuum-v3.9.8.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.8/SHA256SUMS.txt) | +| 3.9.7 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.7) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.7/ComfyUI-H3-Continuum-v3.9.7.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.7/SHA256SUMS.txt) | +| 3.9.6 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.6) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.6/ComfyUI-H3-Continuum-v3.9.6.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.6/SHA256SUMS.txt) | +| 3.9.5 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.5) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/ComfyUI-H3-Continuum-v3.9.5.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.5/SHA256SUMS.txt) | +| 3.9.4 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.4) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/ComfyUI-H3-Continuum-v3.9.4.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.4/SHA256SUMS.txt) | +| 3.9.3 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.3) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/ComfyUI-H3-Continuum-v3.9.3.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.3/SHA256SUMS.txt) | +| 3.9.2 | [Release](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/tag/v3.9.2) | [ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/ComfyUI-H3-Continuum-v3.9.2.zip) | [SHA256SUMS.txt](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases/download/v3.9.2/SHA256SUMS.txt) | + +## 3.9.8 — KJ Model Preview Overrideの継続修正 + +監査済みKJ Model Preview Overrideでは、プレビュー画像を保持しても新Runの保存署名を変えず、別QueueでChunk 1を再利用してChunk 2だけ生成できます。対応は固定したKJソース・Core helperと、**VAE/Audio VAE未接続・Tiny VAE `none`**、Core NoPreviews/Auto/Latent2RGBに限定します。KJNodes 1.5.2全般への対応ではありません。更新後はComfyUIを再起動し、ブラウザーを再読込して新Runを作成してください。旧Take・Workflowの削除は不要です。[対応条件・監査ハッシュ・検証範囲](KJ_PREVIEW_OVERRIDE_V398.md)を参照してください。 + +### 公開版v3.9.8のGPU再検証 — 2026-10-16:PASS + +通常ComfyUIに公開タグ **`v3.9.8`/`58fbd855`** を反映して再起動し、公開版464ファイルとの一致、実ログの3.9.8読込、新しいSampling v6契約内のKJアダプターと固定監査プロファイルを確認しました。新Runを2つ使用した計4 Queueの結果です。 + +| KJ Preview Override併用構成 | 別QueueのChunk 1→2継続 | +| --- | --- | +| Model Attention Backend:PyTorch/Spectrum OFF | **PASS — 1再利用/1生成** | +| Model Attention Backend:Comfy Kitchen INT8+監査済みローカルSpectrum | **PASS — 1再利用/1生成** | + +両構成の2回目は画像を保持したKJ MODELをキャッシュ再利用(**8→16枚**)し、Samplerは実行。契約・revision・元Chunk 1 latentのSHA-256を維持し、`resume_safe=true`でした。映像・音声4本は全デコード成功(512×512/24fps、5秒120フレーム→10秒240フレーム)。保護対象170ファイル(保存Workflow 71件・既存Run/Take 46ファイルを含む)は全ハッシュ不変です。 + +**対応範囲:** 固定したKJ/Core監査プロファイルのみ。ローカルKJソースはupstreamと異なり、KJNodes 1.5.2全般の合格ではありません。Preview OverrideのVAE/Audio VAE未接続・Tiny VAE `none`・Core Auto、RTX 5060 Ti 16 GB、8-step Turbo/Euler/simple、2×5秒。Spectrumは監査済みローカル版のoffline replay、履歴・archiveはsystem RAM、model-aware OFFです。**未確認:** ブラウザーRunボタン/保存再読込、報告者の完全一致graph、別KJ/Spectrum版、SolAttn/preview VAE、カーネル単位の戻り値、主観画質・音質、速度・VRAM比較。保持PNGは32×32で、高解像度プレビュー品質の証明ではありません。既知Inspector問題は残り、旧unsafe/v5 Takeを自動修復・昇格しません。[Issue #31への結果返信](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/31#issuecomment-6209571261)。 + +## 3.9.7 — 外部入力接続時の再生成キャッシュ修正 + +ChunksやVariation Nonceを外部ノードから接続した際に、明示したFull Run再生成がCoreのキャッシュで省略される問題を修正しました。前の採用済みChunk、既存Take、公開入力、公式Workflowは維持します。更新後はComfyUIを再起動し、ブラウザーを再読込してください。Workflowの変更や履歴の削除は不要です。 [Release notes](RELEASE_NOTES_V397.md). + +**公開版3.9.7でのGPU再検証(2026-10-15):PASS。** 公開タグ(`18df80d`)で再起動後、外部Core INTからChunks=2・Nonce=0を入力し、完全に同一のプロンプトでChunk 2を2回再生成しました。自動Nonceは**1→2**へ進み、両回ともChunk 1を再利用してChunk 2だけを生成。以前のTake・raw latentも保持しました。基準生成を含む動画3本は240フレーム・映像/音声各10秒で全デコード成功。既存Workflow 60件・設定1件・Run/Take 30ファイル・公開版459ファイルのハッシュは不変です。条件はRTX 5060 Ti 16 GB、512×512、24fps、2×5秒、8-step Turbo、PyTorch attention/Spectrum OFF、Core API経由。CUDA/OOM/実行エラーは確認されず、ブラウザー・主観画質・他アクセラレーターの受入範囲は拡張していません。 + +元の[3.9.2 Source ZIP](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/archive/refs/tags/v3.9.2.zip)は従来のタグから取得できます。本体ZIPと公式Workflow ZIPは別の配布物です。 + +## 3.9.6 — 再生成指定・Nonce保持と自動Variationの再実行 + +V3.8/V3.9の**Regenerate From・Variation Nonceを保存・再読込後も保持**します。保存ONで**Full Run+明示Chunk+Nonce 0**を指定すると、QueueごとにRun Storageが処理し、完了後の再実行で新Variationを作成できます。前のChunkと既存Takeは保持します。更新後はComfyUIを再起動し、ブラウザーを再読込してください。旧markerなしChunk 1/Nonce 0は、Chunk 1を選び直して一度保存してください。[操作・互換性・検証範囲](REGENERATE_INTENT_V396.md)を参照してください。 + +## 3.9.5 — Issue #31:Attention/Spectrumの保存Take継続 + +CoreのModel Attention Backend(PyTorch/Comfy Kitchen INT8)と監査済みSpectrum Bindingを正確に識別し、新しいRunのReview Each Chunk継続を可能にします。公式Spectrum v0.2.29も専用プロファイルで対応します。以前のunsafe履歴は保持し、再利用可能へ自動変換しません。更新後は新しいRunで生成してください。公開ノード・Workflow・Sampling v6は維持します。[対応実装・更新手順・検証範囲](ISSUE31_SAMPLING_SIGNATURES.md)を参照してください。 + +## 3.9.4 — Registry配布の梱包修正(2026-10-14) + +Manifestに記載したReview復元ガイドを収録し、配布対象外のプロンプト作成資料を除外します。公開前に実際のRegistry ZIPを作り、全収録ファイル・CRC・Manifestの全ハッシュを照合します。生成動作・12公開ノード・公式Workflow・保存済みRun/Takeは変更しません。3.9.3と3.9.2のGitHub Releaseは保持します。 + +## 3.9.3 — 一時ストレージの保存先設定(2026-10-13) + +未圧縮フレームなどの一時データを、指定したストレージへ保存できるようになりました。**設定 → H3Continuum → Storage → 一時ストレージを管理**を開き、**ComfyUIサーバー側**の絶対パスを入力するかストレージ候補を選択して、検証・保存してください。空欄なら従来の保存先を使用します。ワークフローの変更は不要で、V3.9ノードと3.9.1/3.9.2のワークフローを引き続き使用できます。 + +- **FinalizeとDecode Cache Helperの両方**が対象です。指定先の`h3-continuum-scratch`専用領域へ保存します。Run/Takeや完成動画の保存先は別です。 +- 現在・以前の保存先、一時ファイル量、空き容量、使用中/削除可能な容量を表示します。**未使用の一時ファイルを削除**は、参照中の出力・Tensorのview・生存プロセス・所有者不明のファイルを保護します。キャッシュ削除後は再デコード・再書き込みが必要になる場合があります。 +- 保存・削除は実行中と待機中のQueueが空のときに行います。新しいファイルから保存先を切り替え、既存出力は元の場所で保持します。指定先が使えない場合、元のSSDへ黙って戻しません。マルチユーザー運用では共有保存先のUIは閲覧のみです。 +- 使用方式は既存ノードで選びます。**Finalize:Auto/RAM/Disk-backed、Decode Cache Helper:Auto/RAM/Off**。Autoの4 GiB境界とRAM安全余裕は維持します。手動RAMは確保前に空きメモリを、手動Disk-backedは空き容量を確認します。ただし、実行中のメモリ・容量変動やドライブ切断まで防ぐ保証はありません。VRAMを保存領域にする機能は追加していません。 + +**保存先変更は書き込み先を変える機能で、総書き込み量を減らすものではありません。** 物理ストレージが1台なら、同じ機器の別フォルダー・別パーティションへ移しても書き込みを逃がせません。別のSSDは速度を保ちやすく、HDDも使用できますがDecode・結合・保存が遅くなる場合があります。十分なCPU RAMがあればRAM方式で一時書き込みを避けられますが、余裕がなければページファイルへの書き込みやOOMが起こり得ます。CacheをOffにするだけではFinalizeのDisk-backedを無効にしません。容量表示は論理ファイルサイズであり、物理占有量やSSDの累計書き込み量ではありません。対応先はローカルファイルシステムで、ネットワーク共有は正式対応外です。 + +更新時はカスタムノードを更新し、ComfyUIを再起動してブラウザーを強制再読込してください。その後の保存先設定の変更には再起動は不要です。空欄時はFinalizeがComfyUIのtemp、Decode Cacheが既存の`H3_DECODE_CACHE_TEMP_DIR`指定またはOSのtempを使います。フォルダーを指定した場合、その指定がキャッシュ用環境変数より優先します。 + +映像のdtype、動画・音声の結合計算、生成条件の識別は維持します。[実装仕様](TEMPORARY_STORAGE_JA.md)と[検証記録](../PACKAGE_VALIDATION.txt)も参照してください。設定画面のスクリーンショットは[英語README](../README.md)の冒頭にあります。Releaseの公開状態は上表を参照してください。Registry公開は別に確認します。 + +画面にHTTP 404が出る場合は、ComfyUI起動ログを確認してください。旧プロセスが同じアドレス・ポートを占有していると、更新後の起動が失敗し、新しい画面だけが表示される場合があります。残っている旧環境を終了するか別のポートを選んで、新しいComfyUIを起動してください。**Refresh(更新)**は保存先の状態を再取得するボタンで、ComfyUIの再起動ではありません。取得失敗時はHTTP結果と対応方法を表示し、状態を取得できるまで保存・削除を無効にします。 + +## 3.9.2 — Issue #30 Frontend hotfix(2026-10-13) + +ReviewとReference Imagesの不要なUI処理を削減します。 + +- Review比較基準がない場合、上流設定snapshotの作成を省略。 +- Reference選択時のSampler通知を1回へ整理。native値の即時更新と元widget callbackを維持。 +- Referenceノードの目標寸法と実寸が同じ場合、不要なサイズ変更を省略。 + +比較基準がある場合の監視と、Take/Queue時の最新設定確認を維持します。Sampling、保存済みRun/Take、公開Node ID、公式Workflowのbytesは変更しません。引き続きV3.9ノードを使用してください。 + +Frontend修正で**CPU/JavaScript回帰1703件成功・2件スキップ**、独立したInspector再configure問題の**2条件を期待失敗として記録**しました。Coreのオフライン登録・計画確認と隔離Chromium検証も通過しています。このパッケージの実ComfyUIブラウザー確認は[PACKAGE_VALIDATION.txt](../PACKAGE_VALIDATION.txt)に記録します。モバイルでのフリーズ解消は未確認です。Reference表の再構築と、比較基準がある場合のReview走査は残ります。今回のhotfixで新たなGPU生成は行っていません。 + +更新後は**ブラウザーを強制再読込(Ctrl+F5)**してください。表示するパッケージ番号も更新する場合は、通常の手順でComfyUIを再起動してください。3.9.1のRun/Takeは、このUI修正のために移行する必要はありません。固定版は[GitHub Releases](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases)を参照してください。Registry公開は別作業です。 + +## 3.9.1で変わった点(2026-10-04) + +Second Pass、Takeの再利用、Review音声に関するバグを修正しました。 + +- **Second Pass**:音声のみ/複合キーフレームを保持し、途中Reviewや中間Chunk単独でも、その出力区間のconditioningとReference割当を照合して引き継ぎます。 +- **Takeの再利用**:Samplerのclosure、MODELのCFG/wrapperも互換性判定に含め、異なる生成設定のTakeを誤って再利用しないようにしました。 +- **ReviewのDriving Audio**:物理groupの自然時間に合わせて元PCMを1回だけ切り出します。Exact ON/OFFで開始位置を揃えます。 +- **音声resample**:ComfyUI Core標準APIを優先し、古いCoreでは従来方式へ戻します。同じVAE内部の重みを直接変更する場合のReference Encode Cacheの制約も下記に明記しました。 + +**3.9.0から更新する場合**:更新後にComfyUIを再起動し、ブラウザーも更新してください。旧v5 Takeと履歴は読めて、削除されませんが、新しいv6生成の前半として再利用できません。新規Full Video、または **Start again from Chunk 1** で開始してください。公開Node ID・widget順・公式Workflowは維持し、引き続き **V3.9ノード** を使います。 + +修正済みコードは **CPU 1,696成功・1スキップ**、対象を絞ったReview Second PassのGPU検証、生成/Take再利用の統合GPU検証に合格しています。ブラウザー受入は未完了です。任意のLATENT拡大で出る二重輪郭は別課題の [Issue #27](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/27) として管理し、今回の修正の必須Gateには含めません。今回もmainソースの更新で、GitHub Release/Registry公開ではありません。 + +![H3 Continuum V3.9:固定の参照画像番号、チャンク別割り当て、Samplerへの1本の接続](images/v39/v39-feature-summary.png) + +*現在の`main`はV3.9です。Image 1~9の固定番号、チャンク別のReference選択、V3.9 Samplerへの1本の接続が新しい操作です。* + +## 3.9.1の修正詳細 + +Second Passで音声のみのキーフレームを保持し、複合キーフレームでは映像だけを空間適応します。ReviewのDriving Audioは、区間選択前の物理decode groupの自然時間から元PCMを切り出します。Exact ON/OFFの開始位置は共通です。完全に空のReview音声だけは選択映像の尺に合う無音を補い、部分的に短い音声はExact OFFで引き延ばしません。 + +新しい生成はSampling Contract v6/Graph Contract v4を使用します。Samplerのコード・closureとMODELのCFG/wrapper等の設定も再利用判定へ含めます。旧v5のTake・raw・生成履歴は読めますが、v6生成へ継ぎ足せません。新規Full Video、またはReviewの「Start again from Chunk 1」で新しいrevisionを作成してください。旧Takeは残ります。署名を観測できない外部設定でも新規生成は可能ですが、自動再利用は無効になります。明示した不適合TakeやRetryを別Takeへ置き換えません。 + +Reference/DrivingのresampleはCore標準APIを優先し、そのAPIがない旧Coreだけ従来のTorchAudioを使います。最終出力へ渡す元PCMの波形とsample rateは保持します。 + +**Reference Encode Cacheの制約**:同じVAEオブジェクトの内部重みを直接書き換えた場合、その変更を自動検知しません。変更後は`v3.ref_encode_cache.clear_ref_encode_cache()`を呼ぶか、新しいVAEオブジェクトへ再ロードしてください。通常の同一VAE・同一入力ではcache HITを維持します。Decode Cache Helperのresetは別のcacheを対象とし、Reference Encode Cacheを消しません。 + +この修正はV3.9内の変更です。公開Node ID・socket・widget順、公式Workflow、Run Storage v3、State/Session形式を維持し、LoRA Planは追加しません。CPU検証と実ブラウザー・GPU受入は別です。 + +V3.9では、Second Passへ渡す条件を実際のReview出力区間へ対応させます。途中までの出力や中間Chunk単独でも、条件の照合に成功すれば、その区間自身のconditioningとReference割当を継承します。全体の完了状態や保存済みTakeは変更しません。条件が不足・不一致の場合は、従来の診断付きprompt-only fallbackを維持します。 + +修正済みランタイムは**CPU 1696成功・1スキップ**、**512×512・24fpsのGPU 3ケース**(途中2区間、Chunk 2単独、完了済み3区間の再利用)で合格しました。全区間の条件継承を確認し、保存FLACと元PCMの対応区間は完全一致、既存Takeとrawファイルも保持しています。合格対象は修正対象の動作です。ブラウザー受入は未完了で、任意のLATENT拡大時の二重輪郭はIssue #27で別途調査します。過去のV3.8 Release/tagはそのまま利用できます。 + +## V3.9の基本操作を動画で見る + +[H3 Continuum V3.9 — Getting Started: Chunks, Prompts & Reference Images](https://youtu.be/AKJxBXaiG6Q)(12分26秒) + +First Imageの基本設定、チャンクの設定、Review/Takeでの再試行、Timeline形式のプロンプト、Reference Imagesの使い方を紹介します。英語のナレーションと字幕、7つのチャプターがあり、バージョンを明記したUIの静止画と実際の生成例で説明します。 + +![公開当時のH3 Continuum V3.8X2機能紹介画像](images/v39/v38x2-feature-summary-historical.png) + +*V3.8X2の画像も比較用に残しています。画像内の「CURRENT MAIN」は作成当時の表記で、現在の`main`はV3.9です。V3.8 SamplerとV3.8X2 Workflowは利用できますが、Reference配線はV3.9へ自動変換されません。* + +## main更新:外部INTによるWidth/Height入力の修正(2026-10-01) + +V3.8/V3.9のWidth/Height入力と表示コントロールの対応を修正しました。外部INTの接続位置をWorkflow保存・再読込やタブ切替後も維持します。接続中の欄は編集不可になり、表示する数値は保存済みのManual予備値です。上流ノードが実行時に出す値を推測して表示するものではありません。接続を外すと手動編集に戻ります。 + +Coreの **Int** ノードから正確な幅・高さを指定する場合は、**Size Source = Manual**にしてWidth/HeightのINT入力へ接続してください。**First Image**は引き続き画像とResolution設定からサイズを決めます。INTを接続しても優先順位やSize Sourceは自動で変わりません。有効なFirst Imageがない場合は従来のManualフォールバックを使います。寸法検証、Backend入力順序・初期値は変更していません。 + +確認済み:**全CPU試験1569 PASS/1 skipped**(新Frontend試験22/22・既存Review試験60/60を含む)、両バージョンのChrome保存・再読込とタブ5往復。外部INTを使ったV3.9のGPUテスト1回では、**480×640・24fps・96フレーム・映像/音声とも4秒**を正確に出力しました。20ステップ、全体**107.3秒**で、保存動画の全フレームを正常にデコードできました。短い単一条件の確認であり、長時間継続や全モデルでの受入確認ではありません。Sampling、Run Storage、先行するReview/キャッシュ修正、公式Workflow JSON/ZIP、過去Releaseは変更していません。 + +Continuumを`main`から更新し、**ブラウザーをCtrl+F5で強制更新**してください。このJavaScript修正だけならComfyUI本体の再起動は不要です。同時にPython修正や新ノード定義も更新された場合は通常どおり再起動してください。今回もmainのソース修正で、新Release/Registry公開ではありません。 + +## main更新:V3.9の条件付け準備の速度修正(2026-10-01) + +**Prompt Format = Fixedかつ有効なReference Imagesがない場合**に、V3.9で余分なプロンプト条件付けの準備が繰り返される経路を修正しました。入力が同じ場合に既存の上限付きconditioningキャッシュを再利用します。First Imageは利用できます。List/Timeline、有効なReference Images、Reference Audio、Timeline Video、終端ペアのgroupは従来の経路を維持します。Decode Cache Helperとは別の修正で、Samplingは省略しません。 + +検証した`2 × 5秒`・480×640・First Imageの条件では、group準備が約**10.76秒から0.013秒**へ短縮しました。修正後の同セッション比較(warm-up後、各版3回)では、全体時間の中央値は**V3.8が273.58秒、V3.9が273.78秒**で測定範囲も重なり、以前の約6%差は見られなくなりました。全Workflow/モデル/GPUでの解決や、GPU演算そのものの高速化を確認したという意味ではありません。 + +Continuumを`main`から更新し、Python修正を読み込むため**ComfyUIを完全に再起動**してください。確認済み:**CPU 1558 PASS/1 skipped、GPU生成8回すべて成功**(ComfyUI 0.38.0/RTX 5060 Ti 16 GB)。Seed、チャンク/フレーム契約、Run Storage、V3.8X2/V3.9公式Workflow・ZIP、過去Releaseは変更していません。今回もmainのソース修正で、新Release/Registry公開ではありません。 + +## main更新:Review表示の復元修正(2026-10-01) + +別のWorkflowタブへ切り替えて戻ると、保存済みTakeが残っていてもReview操作が消える問題を修正しました。Workflowの復元完了後に履歴を取得し、古い画面の非同期結果は反映しません。実際に生成設定を変えた場合のReview/Take適用ガードは維持し、元の設定に戻すと操作が復帰します。 + +カスタムノードを`main`から更新後、ブラウザーを更新して新しいJavaScriptを読み込んでください。確認済み:Frontend回帰60/60、関連CPU試験108件、Chromeのタブ5往復とBase Seed変更/復元。Take管理データは不変です。Sampling、Run Storage契約、V3.8X2/V3.9のWorkflowファイルと過去Releaseは変更していません。この修正のGPU継続生成は未検証です。[詳細と検証範囲](REVIEW_RESTORATION_REPAIR.md)。 + +## アーカイブされた実装・受入記録(V3.8現行ガイドではありません) + +以下は一時的にsource historyとして残る旧リリース記録です。V3.8の標準経路・公開範囲・推奨配線を示すものではありません。旧WorkflowにはMigration Policyと対応するhistorical Release/tagを使用してください。 + +### V3.7 高解像度Refinement基盤 + +V3.7では`H3 Continuum Sampler V3.7`を追加し、解像度を変更するSecond Passに必要な2つの基盤を完成させました。V3.6のProduction初期値と既存SIGMAS socketは変更していません。 + +### 解像度変更後もConditioningを再構築 + +Conditioning Adapterは、First/Last Image、Continuation context、Still Image Guideの元画像をtarget latent geometryに合わせて再構築します。Physical group ownership、Long Terminal Merge、Core側のPackedLayout/RoPE再構築、First Pass Audioのpassthroughは維持します。704×704と1152×1152のCorrectness GateはPASSしていますが、大きなcanvasほどmemory使用量と処理時間は増えます。 + +### RefineScheduleでRefinement範囲を明示 + +内部のversioned contractである`RefineSchedule`は、`External`、`Full`、`Tail`、`Partial`に対応します。`External`は入力SIGMAS tensorをそのまま維持し、`Tail`と`Partial`は同じsource scheduleから範囲を正確に切り出します。内部Tail 6 Gateでは、6 evaluationのsuffix再現とdecoded Audio PCMのbit-exact passthroughを確認しました。現時点では実装基盤として内部で扱い、公開widgetは追加していません。Production初期値も変更していません。 + +### Still Image GuideはExperimental + +V3.7では、Still Image Guide 1枚をabsolute frameからowner physical groupへ割り当て、高解像度Second Passでも元画像から再encodeできます。位置変換、Terminal Merge ownership、Run Storage identity、non-owner groupの非干渉はCorrectness PASSです。一方、Core Add Guideはhard anchorとして働くため、anchor位置で急なtrajectory変化が起き、その後のmotionも別の軌道へ分岐する場合があります。**Still Image GuideはExperimentalで、Production昇格はHOLDです。** 滑らかなtransition制御としては扱わないでください。 + +### V3.6.1 Maintenance Hotfix + +V3.6.1は、受入済みBalanced 22 Masked AV経路を変更せず、限定的なfail-open/互換性問題を修正します。Timeline/List記法が混在したPromptは`H3C-P105`を表示し、Timeline解析ではstandalone `---`行だけを除去します。未指定Chunkの既存fallbackとFixed Promptのfail-open動作は維持し、Prompt構文だけを理由に生成を停止しません。 + +`Continuation Backend = Standard`かつAudio Continuity ONでは、`Balanced — 22 frames`が引き続きMasked AVを使用します。`Fast — 5 frames`、`Strong — 39 frames`、`Auto — conservative`はRun Storage identity作成前に、受入済みReference Contextへ解決します。UIの選択値は書き換えず、status reportへ解決結果と理由を表示します。Audio Continuity OFFのStandardはMasked Video、Compatibilityは常にReference Contextのままです。 + +Run Storageは、Last Frame未接続を表す値をすべて同じ空identityとして扱います。通常のLast Frameなし2チャンク生成を3チャンクへ延長すると、旧最終Chunkが`none`だったV3.6.1 cacheを含めて`2 reused, 1 generated`になります。実Last Frame接続時とLong Terminal Mergeのatomic pairは従来どおり厳密です。Run StorageをOffへ変更した場合は、Queue前にhiddenのRegenerate FromとVariation Nonceも`Auto`と`0`へ戻します。 + +### V3.6 Masked AV Continuation + +V3.6では`H3 Continuum Sampler V3.6`を追加しました。標準backendは、前Chunkで確定したVideo/Audio latent prefixを次のH3 Target内へ直接配置し、CoreのNoise Maskで保持します。V3.5のReference Context方式のように、同じ過去フレームを別Reference blockとして追加しないため、H3が処理するpacked sequenceを短縮できます。 + +| Continuation Backend | 用途 | +|---|---| +| `Standard` | 推奨V3.6経路。Balanced 22+Audio ContinuityはMasked AV、Fast 5/Strong 39/Autoは安全にReference Contextへfallbackします。Audio Continuity OFF時はMasked Videoです。 | +| `Compatibility` | 受入済みV3.5 Reference Context動作へ戻すAdvanced fallbackです。 | + +内部transport識別子はUIへ表示しません。Run Storage identityは実行前に解決したtransportへ従います。Balanced 22 Masked AVはReference Contextと分離し、非Balanced Standard fallbackは同一動作のReference contractを安全に共有します。`Regenerate From`でもLong Terminal Mergeのlogical `[2,3]`を1つのphysical sampleとして再利用・再生成します。 + +### 受入結果 + +- T2VA、I2VA、Reference Image+Reference Audio、Balanced 22-frame Joint AV、3×5秒FL2VA Long Terminal MergeのGPU GateをPASS。 +- 保護Video/Audio prefixは最終的にbit-exact。生成領域はSamplerの出力を維持します。 +- Reference Image+Reference Audio代表出力は実聴PASSで、click、dropout、定位異常の報告なし。 +- 実測したFL2VA Terminal Group 2では、StandardがCompatibilityよりpacked rowsを2,874行(`8.01%`)削減し、Sampling中央値を`4.06%`短縮。速度はmodel、hardware、workflowで変わります。 +- `chunk_seconds`は4.0~30.0秒、初期値5.0秒、step 0.1です。5~15秒を推奨・検証済み範囲として維持し、長時間・高解像度ChunkはVRAMと処理時間が大きく増える場合があります。 + +V3.5.3、V3.5、V3.4のNode IDと保存済みWorkflowは維持します。V3.5.3は従来Reference Contextのままで、V3.6標準backendへ暗黙に切り替えません。 + +## V3.5.3 Maintenance Hotfix + +V3.5.3は配布整合性だけを修正するメンテナンス版です。公開Conditioning Bridge WorkflowのWidth/Height接続、V3.4/V3.5テンプレート内の孤立link ID、Hybridの``警告番号、厳密な画像decoderで読めなかった旧PNG 1件を修正しました。 + +生成動作、Node、socket、Sampling、Conditioning payload、Terminal Merge、Assembly、Seam、Run Storage、Prompt/CLIP cache、Video Guide最適化、V3.4互換性は変更していません。既存V3.5.x Workflowはそのまま読み込めます。 + +## V3.5.2 Stabilization & Optimization Update + +V3.5.2は新しい生成モードを追加する版ではありません。V3.5.xを安定化し、安全な条件で繰り返しPrompt/CLIP処理を省略し、長いVideo Guide入力の一時メモリを削減します。既存Workflowと生成契約は維持します。 + +![H3 Continuum V3.5.2の安定化・最適化結果](images/v352-stabilization-optimization.png) + +画像は受入済み最適化の実測値です。最終パッケージゲートでは、さらにComfyUI 0.33.3と、この画像を含む172項目のManifestでPASSを確認しています。 + +### 主な変更 + +1. **繰り返し実行Prompt/CLIP cache**: 条件が変わらないT2VA、I2VA、FL2VAではCPU上のPrompt/CLIP結果を再利用できます。Reference、schedule、tokenizer option、hook変更CLIPは安全側でbypassします。Sampling自体は通常どおり再実行します。 +2. **Video Guideメモリ最適化**: source全体をfinite検査とSHA-256 identityへ含めたまま、VAE/conditioningに必要なprefixだけを保持します。 +3. **安定化監査**: Sampling、Decode、Assembly、保存、任意機能の負荷、保持memoryを実測しました。根拠のないSampling、Driving Audio、Audio hash、Assembly、Seam、Session、V3.4最適化は採用していません。 + +### 実測結果 + +| 項目 | 変更前 | V3.5.2 | 結果 | +|---|---:|---:|---:| +| T2VA繰り返しPrompt/CLIP | 5.898398秒 | 0.000028秒 | Cache HIT、`encode_calls=0` | +| FL2VA繰り返しPrompt/CLIP | 21.642687秒 | 0.007128秒 | Cache HIT、`encode_calls=0` | +| Video Guide Peak追加RSS | 396.8 MiB | 114.7 MiB | 約71%削減 | +| Video Guide保持Storage | 225 MiB | 93 MiB | 約59%削減 | + +Prompt/CLIPの数値はconditioning区間だけで、総生成時間ではありません。Video Guideメモリ比較は300×256×256 RGB入力/必要prefix 124 framesの固定CPU条件です。別のGPU A/Bでは実際の15秒/360-frame Video Guideを使い、decoded videoとaudio PCMの完全一致を確認しています。 + +RTX 5060 Ti 16 GB/RAM 64 GBの検証環境で測定したSage-only Production baselineは、576×576 T2VA 1×5秒が168.069秒、640×640 FL2VA Long Terminal Merge 3×5秒が379.765秒です。環境・設定固有の測定値であり、すべての環境に対する速度保証ではありません。Samplingが最大コストで、Continuum Assemble + Seamは1%未満でした。 + +**V3.8.0はhistorical release baselineです。V3.8X2はpackage 3.8.3の別Workflowとして維持し、V3.9はmainの現行ソースです。** V3.8X2が内部利用する旧module/classはsourceへ維持します。現在の登録面はV3.8X2の10 IDとV3.9追加の2 IDです。一部は旧IDを維持しますが、それ以外のIDを必要とする旧保存Workflowは対応するhistorical Release/tagを使用してください。Still Image Guideは引き続きExperimentalです。 + +## V3.5.1 Reference Audio/互換性更新 + +V3.5.1ではV3.4のSampling、Conditioning、Terminal Merge、Assembly、Seam、Run Storage契約を変更せず、次の2点を追加しました。 + +1. **Reference Audio(任意)**: 生成音声を置換しない、H3 nativeの音声conditioningです。Workflowの保存・再読込互換を守るため、2つのsocketは常時表示します。 +2. **Conditioning Bridge V3.5**: Continuumのphysical groupごとに、完全なCore互換`MODEL`と`CONDITIONING`を外部Samplerへ公開します。 + +Reference Audio socketはPython `INPUT_TYPES`で常設し、保存・再読込時の値ずれを防ぐため動的socket変更とUI-onlyの`Hidden`/`Show` controlを削除しました。V3.4のNode IDとbackend socket keyは保存済みWorkflow互換のため維持しており、既存V3.4/V3.5 Workflowの接続は変わりません。 + +### Reference系入力の違い + +表示名は役割の違いを示しています。名前が似ていても同じ組み合わせではありません。 + +| 表示入力 | 接続元 | 用途 | 最終音声 | +|---|---|---|---| +| `reference_image_1`~`reference_image_3` | 静止画`IMAGE` | 人物、被写体、外観などの継続参照 | 音声なし | +| `Video Guide Frames` | Video loaderの`IMAGE`フレームbatch | 全chunkでmotion、framing、timing、appearanceをガイド | 元動画の音声は含まない | +| `Driving Audio` + `Driving Audio VAE` | Audio loader、またはVideo loaderの`AUDIO`と対応Audio VAE | 音声で生成をガイドし、入力音声を最終出力へ維持 | 生成音声を選択した元音声へ置換 | +| `Reference Audio (Optional)` + `Reference Audio VAE (Optional)` | 単独Audioと対応Audio VAE | H3 native conditioning専用 | 生成音声をcopy・置換しない | + +音声付き動画を使う通常の接続は、loaderの`IMAGE`を`Video Guide Frames`へ、`AUDIO`を`Driving Audio`へ接続します。conditioning専用動作が目的でない限り、動画の音声を`Reference Audio (Optional)`へ接続しないでください。 + +![Sampler V3.5の常設Reference Audio、Video Guide Frames、Driving Audio、Reference Image入力](images/v351-video-guide-frames.png) + +`Reference Audio (Optional)`と`Reference Audio VAE (Optional)`は、Pythonのinput schemaどおり常時表示します。V3.5.1ではこの2 socketをFrontendで動的に追加・削除せず、UI-onlyの`Hidden`/`Show` widgetも使用しません。これによりWorkflowを保存・再読込した際のwidget値の位置ずれを防ぎます。backend input keyと既存接続は変更しません。 + +![常設Reference Audio socketと正常なwidget配置](images/v351-reference-audio-permanent.png) + +### Conditioning Bridge V3.5 + +`H3 Continuum Conditioning Bridge V3.5`は外部sampling用のAdvanced接続点です。`model`、`clip`、外部処理後の`video_latents`、`assembly_plan`、`refine_context`を接続し、選択したconditioning経路で必要な場合だけ`video_vae`を接続します。出力`group_models`と`conditioning`はphysical group数と同数で、各要素は完全なComfyUI objectのままです。CONDITIONING内部entryをphysical-group listへflattenしません。 + +```text +H3 Continuum Sampler V3.5.video_latents + → LBH等の外部H3 latent処理 + → H3 Continuum Conditioning Bridge V3.5 + → Core BasicGuider + 外部Sampler + → Core Video / Audio VAE Decode + → H3 Continuum Assemble + Seam V3.5 +``` + +AV LATENTの組、Noise、SIGMAS、Audio Lock、sampling、audio passthroughは外部workflow側の責任です。Bridgeは準備済み`MODEL`と`CONDITIONING`、更新済みAssembly Planだけを公開します。 + +![V3.5.1 LBH+Conditioning Bridge外部sampling接続図](images/v351-lbh-conditioning-bridge-flow.svg) + +退役したV3.5.1 LBH+Conditioning Bridgeの接続例は、対応するhistorical Release/tagでのみ参照できます。これはV3.8 Workflowではありません。 + +接続例ではCore標準名の`BasicGuider`、`BasicScheduler`、`SamplerCustomAdvanced`を変更していません。公開サンプルではCoreノードを独自タイトルへ変更せず、Continuumノードや外部カスタムノードと直感的に区別できる状態を維持します。LBH latent upscaler、AV LATENTの結合/分離、load/save、任意accelerationは外部ノードなので、各ComfyUI環境に合わせて導入または置換してください。 + +LBHはlatent geometryだけを変更し、Continuumのdenoise強度は持ちません。拡大後latentをどれだけ再生成するかは外部`BasicScheduler`のSIGMASで調整します。1.5x等へ上げても高速な場合がありますが、target canvasに応じてmemory、decode、sampling負荷は増加します。 + +## V3.5の主な追加機能 + +V3.5の大きな追加は2点です。 + +1. **Continuum対応Second Pass / Hi-Res Fix**: physical group構造を維持した再Samplingと、標準Pixel/VAE 2x経路。 +2. **低メモリAssemble + Seam**: Exact Duration、Seam、Terminal Merge、Audioを維持したRAM / Disk-backed / Auto出力。 + +### 1. Second Pass / Hi-Res Fix + +`H3 Continuum Hi-Res Fix V3.5`はVideo VAE Decode → pixel resize → Encode → low-denoise Second Passを1ノード化したMain経路です。長尺2xではVRAM上限に達する場合があるためExperimentalです。 + +![H3 Continuum Hi-Res Fix V3.5 node](images/v35-hires-fix-node.png) + +通常のHi-Res Fix接続: + +```text +H3 Continuum Sampler V3.5 + → H3 Continuum Hi-Res Fix V3.5 + → Core Video / Audio VAE Decode + → H3 Continuum Assemble + Seam V3.5 +``` + +`H3 Continuum Second Pass V3.5`は外部加工したH3 video latentをphysical group構造のまま再Samplingする安定版Advanced入口です。Issue #8への本体回答です。 + +![H3 Continuum Second Pass V3.5 node](images/v35-second-pass-node.png) + +外部latent upscaler接続: + +```text +H3 Continuum Sampler V3.5.video_latents + → LBH等の外部H3 latent処理 + → H3 Continuum Second Pass V3.5 + → Core Video / Audio VAE Decode + → H3 Continuum Assemble + Seam V3.5 +``` + +画像では`BasicScheduler`を省略していますが、Second Pass用`SIGMAS`の接続は必要です。GPU受入済み基準は`res_multistep`、`simple / 10 steps / denoise 0.35`、Lanczos pixel resize、FL2VA 1×5秒・576→1152です。Hybrid FL2VA + Referenceも1×5秒で、統合2x Hi-Res FixとAdvanced Second Pass直結の両方を受入しました。 + +Hi-Res Fixを接続しなければV3.4 Sampling経路は変わりません。強制CPU offloadや低VRAM samplerは追加していません。`enabled=false`ではResize、VAE、Second Passを評価せず、元のVideo/Audio LATENTとAssembly Planをそのまま返します。 + +### 2. 低メモリAssemble + Seam + +`H3 Continuum Assemble + Seam V3.5`はRAM / Disk-backed / Autoを選択できます。Disk-backedでは完成Video IMAGEをWindows対応mapped fileへ直接writeし、AudioはRAMに維持します。 + +1536×1536・360 frames・完成IMAGE 9.49 GiBの実測では、RAM版とDisk-backed版の映像/音声hashが完全一致しました。 + +| Windows process指標 | RAM | Disk-backed | 削減 | +|---|---:|---:|---:| +| Private memory | 12.41 GiB | 2.91 GiB | **9.50 GiB** | +| USS | 10.29 GiB | 0.80 GiB | **9.49 GiB** | + +これはContinuum Assembly時の**システムRAM/private commitment削減**であり、SamplerのGPU VRAM削減値ではありません。Disk I/Oにより最終Assemblyが遅くなる可能性はありますが、前段のモデルSampling速度は変えません。 + +### GPU受入範囲 + +| 経路 | 条件 | 結果 | +|---|---|---| +| Main Hi-Res Fix | FL2VA 1×5秒、576→1152 | PASS | +| Main Hi-Res Fix | Hybrid FL2VA + Reference 1×5秒、576→1152 | PASS | +| Advanced Second Pass | Hybrid/Reference 1×5秒、576×576 | PASS | +| Main Hi-Res Fix | FL2VA 3×5秒、576→1152、RTX 5060 Ti 16 GiB | **Terminal 77TでCUDA OOM** | + +3×5秒2xではFirst Passと37T groupのSecond Passは完了し、最後のlogical `[2,3]`を維持した77T physical groupが1152×1152の最初の推論でVRAM上限に達しました。group構造の破損ではなく、16 GiB実機でのリソース上限です。 + +詳しいsocket接続・OFF時のpassthrough・既知の制限は[英語版V3.5接続ガイド](V35_HIRES_FIX.md)を参照してください。 + +## 公開テンプレートワークフロー + +- [V3.9 Workflow JSON](../examples/workflows/MiniMax_H3_Continuum_V39.json)/[単独ZIP](../examples/workflows/MiniMax_H3_Continuum_V39.zip) — 現行V3.9 graph。ZIPにはこのJSONだけを収録 +- [V3.8X2 Workflow JSON](../examples/workflows/MiniMax_H3_Continuum_V38X2.json) — 併存する旧系公式graph。Spectrumは初期状態でOFF +- [V3.8X2 Workflow ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2.zip) — V3.8X2とV3.9の両JSONを格納。カスタムノードのインストーラーではありません +- [V3.8X2 Helper明示名JSON](../examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.json)/[ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.zip) — 同じgraphを説明的な配布名で収録 + +Registry配布予定の対象はV3.9とV3.8X2のJSON/ZIPです。依存なしのテンプレートではなく、Spectrum・rgthree・KJNodesが必要です。ComfyUI-Easy-Useは不要です。Turboへ切り替えてもgraph内の外部ノードは残ります。SpectrumをOFFにし、用途に合うLightX2V Turbo LoRAを1本だけONにしてsampler/Stepsを合わせます。保存状態はSpectrum OFF、Turbo LoRA OFF、res_multistep/simple/20 Stepsです。 + +配布元のprompt、画像・音声名、ノード表示名、設定は変更していません。手元にあるファイルを選び、不要な入力をOFFにし、promptを入力してください。モデルやメディアは同梱しません。`Save 3x5s Video`などの保存済み表示名は生成時間を決めません。実際の長さはSamplerのChunksとSeconds per Chunkで決まります。 + +旧Workflowは開発履歴としてsource repositoryへ保持しますが、V3.8 Registry packageからは除外します。旧保存Workflowは[Migration Policy](V38_RELEASE_AND_MIGRATION.md)に従い、対応するhistorical Release/tagで開いてください。Registryの正式pack検証・公開はGitHub sourceの更新とは別工程です。 + +## V3.4互換 + +V3.4の実装moduleは内部継承とhistorical testのためsourceへ保持しています。現行の登録面は上記12 IDに限定します。それ以外のIDを必要とするV3.4保存Workflowは、対応するhistorical packageを使用してください。 + +## V3.4.0 Stable + +V3.4.0は、最大3枚のReference Image、Run Storage、Spectrum Interop、Decode後のAudio / Video Seam補正を維持し、`Driving Audio`と`Video Guide Frames`を正式な入力として追加します。 + +`Driving Audio`は入力音声を絶対時間で各Chunkのガイドに使い、最終出力には元音声を維持します。`Video Guide Frames`はVideo loaderが出力したIMAGE frame batchを全Chunkで継続使用し、`Efficient - 0.4 MP`、`Balanced - 0.6 MP`、`Match Output`から処理解像度を選択できます。青いIMAGE socketですが、通常の1枚の静止画参照ではなく動画フレーム列を接続します。 + +### Size項目の条件付き表示と内部Resize + +V3.5のCompact UIでは、対応する入力を接続した時だけSize項目を表示します。 + +- `reference_image_1`~`reference_image_3`のいずれかを接続すると、`Reference Size`を表示します。 +- `Video Guide Frames`を接続すると、`Video Guide Size`を表示します。以前のWorkflowやIssueで`Video Reference Size`と記載されている項目と同じもので、V3.5.1では表示名だけを`Video Guide Size`へ明確化し、既存のbackend keyと保存済みWorkflowの互換性を維持しています。 + +したがって、新規ノードで見えないSize項目は欠落ではなく、未接続のため非表示になっています。Reference ImageとVideo Guide FramesはVAE Encode前に内部Resizeされます。Video Guide Framesはアスペクト比を維持してH3 canvasへ整列し、縮小が必要な場合はLanczosを使用します。小さい入力を自動的に拡大はしません。通常は外部Resizeノードを必要としませんが、意図的なcrop、小さい素材の強制upscale、独自のresize/upscale方式を使う場合には外部処理を利用できます。動画のdecodeとframe-rate変換は引き続きloader側の責任です。 + +Reference入力のバイパスされたソケットは無視され、有効な画像だけがPicture 1から連続採番されます。空欄や不完全なプロンプトも停止せず、警告とFixed fallbackで生成を続行します。未知のモデル、ノード、プロンプト形式を独自判断で拒否する制限も撤廃しました。 + +## Run Storage:第2部分だけを再生成 + +通常のT2VA/I2VAを`2 chunks x 5 seconds`にした場合、Chunk 1が最初の5秒、Chunk 2が次の5秒です。 + +初回生成前に次を設定します。 + +1. `Run Storage`を`Save + Auto Resume`にします。 +2. 固定の`Run Name`を選び、Base Seedも固定します。 +3. まず10秒全体を一度生成します。Continuumは完了した各Chunkのraw Video/Audio latentを保存します。 + +最初の5秒を保持して、第2部分だけを作り直す手順は次のとおりです。 + +1. Run Name、Base Seed、model、LoRA、解像度、sampler、SIGMAS/steps、Reference、Continuity設定を初回と同じに保ちます。 +2. `Regenerate From`を`Chunk 2`にします。 +3. `Variation Nonce = 0`なら、完了済みの再生成に対して新しいバリエーションを自動選択します。互換性のある未完了runを再開する場合は、既存のバリエーションを引き継ぎます。`1以上`は指定値を固定するため、同じ入力で別の結果を求める場合は別の正の値へ変更します。Base Seedは固定のままにします。 +4. Workflowをもう一度Queueします。 + +Review画面の`Try this chunk again`は、この手動指定とは別に、内部のバリエーションを自動で進めます。 + +Continuumは保存済みChunk 1を再Samplingせずに読み込み、その確定済み末尾をContinuation contextとして新しいChunk 2をSamplingします。Reportの目安は`1 reused, 1 generated`です。その後、10秒全体を再Decode/Assemblyします。第1ChunkのSampling結果は再利用され、境界のSeam処理は新しいAssembly時に再評価されます。 + +第1部分と第2部分で異なる指示を使う場合は、初回生成前にListまたはTimeline Promptとして分けておきます。保存後にPrompt Planや生成Contractの入力を変更すると新しいRevisionになり、以前のChunk 1を再利用できない場合があります。初回を`Run Storage = Off`で生成した場合も、後から遡って再利用することはできません。`Regenerate From`はChunk境界単位であり、1つのChunk内部の任意領域を編集する機能ではありません。 + +FL2VA Long Terminal Mergeは重要な例外です。最後の2つのlogical Chunkを1つのatomicなphysical Samplingとして扱うため、その2つは一緒に再生成されます。 + +Run Storageの保存契約にはDriving AudioとVideo Guide Framesのidentityを含め、入力変更後に古いChunkを誤再利用しないようにしています。 + +## Driving Audio / Video Guide Frames / Seam + +Driving Audioだけを使う、Video Guide Framesだけを使う、Video Guide Framesと別音声を組み合わせる、という3通りの接続に対応します。Video Guide Framesは24 fps入力を基準とし、異なるfpsの素材はLoad Video側で24 fpsへ設定してください。 + +H3 Continuum Assemble + SeamはAudio Seam AutoとVideo Seam Autoを既定とし、フレーム削除を行わずに境界の一時的な露出・色変化を補正します。Auto 2は露出ランプ向けの実験的な追加モードです。 + +## Legacy V2.1.7 Stable Facade UI + +V2.1.7では、V2.1.6の動的表示制御を廃止し、標準ComfyUI描画だけを使用する静的Facadeへ移行しました。生成処理は互換用`H3ContinuumSamplerV2`を再利用し、State、Session、Prompt Plan、Spectrum Interop、保存Schemaは変更していません。 + +通常表示は次を中心にします。 + +```text +model / clip / video_vae / audio_vae / sampler / sigmas +Sequence Prompt +first_frame +prompt_overrides(任意pack) +advanced(任意pack) + +Prompt Format +chunks +chunk_seconds +width / height +continuity +base_seed +control after generate +Seam Correction +Seam Correction + +outputs: images / audio / result +``` + +### 補助ノード + +- `H3 Continuum Clip Overrides`: Clip別Promptを1本のpackへまとめます。 +- `H3 Continuum Advanced`: resume/session入力と低頻度設定をまとめます。 +- `H3 Continuum Result`: resultから`last_state / session / report`を展開します。 +- 補助ノードを使わない通常生成では、Sampler本体だけで動作します。 + +### Legacy互換 + +当時のV2.1.7では、旧`H3ContinuumSamplerV2`をLegacy/Coreノードとして登録し、既存WorkflowのノードID、入力順、出力順を維持していました。これはV3.8の公開範囲を示す説明ではありません。このIDを使う旧Workflowは対応するhistorical Release/tagで開いてください。 + +## Continuity + +- `Auto — conservative` +- `Balanced — 22 frames`(標準) +- `Fast — 5 frames` +- `Strong — 39 frames (Experimental)` + +5/22/39フレームはH3の時間latent周期`1,4,4,4,4`に対して、それぞれ2/7/12 temporal latent stepsです。 + +## Spectrum + +Continuumは有効な前チャンクContextがある場合だけ`h3_continuum` Interop API v1をMODEL optionsへ付与し、対応Spectrumへ`min_actual_prefix_steps=2`を要求します。初回チャンクにはInterop keyを付与しません。未知・未対応Spectrumでは通常Spectrumとしてfail-openします。 + +## 推奨配線 + +```text +UNET -> SageAttention -> Sol-Attn -> LoRA -> Spectrum -> H3 Continuum Sampler +``` + +まず`chunks=2 / chunk_seconds=5 / Balanced 22f / Seam Correction=Off`でNative Continuityを確認し、その後`Seam Correction=Auto`を比較してください。 + +## インストール + +V3.8はComfyUI 0.34.2で検証しています。以下の歴史的な受入記録には旧ComfyUI版も登場しますが、現在のインストール目標ではありません。V3.8の依存を満たす限り、ComfyUIの最新版は必須ではありません。 + +ZIPを展開して、`ComfyUI-H3-Continuum`フォルダーを`ComfyUI/custom_nodes/`へ置き、ComfyUIを再起動します。 + +再起動後はV3.8X2 Workflowで`H3 Continuum Sampler V3.8`、V3.9 Workflowで`H3 Continuum Sampler V3.9`と`H3 Continuum Reference Images V3.9`を確認してください。現行の登録面は計12 IDです。 + +旧`ComfyUI-H3-Continuum-Join`が残っている場合は同時ロードを避けるため削除または退避してください。同梱`install_windows.bat`は旧名・新名の既存フォルダーを日時付きでバックアップします。 + +## 検査 + +Public Surface suiteはV3.8X2の10 IDとV3.9を含む計12 ID、公式Workflowと各ZIP内JSONの同一性、外部依存、Registry除外、既存V3.8 widget/socket順、`Show Advanced Settings`/`Hide Advanced Settings`による表示切替を確認します。Registry配布対象のハッシュは`REGISTRY_MANIFEST.sha256`、source側の整合性は`MANIFEST.sha256`で管理します。 + +**過去のV3.8X2準備証拠:** Full CPU suiteは**1,367 passed / 1 skipped / 0 failed**です。従来の最終GPU Functional Gateでは`3 × 5秒`のReview Each Chunkを実行し、Q1~Q3はphysical groupを1つずつ生成、Q4は`3 reused / 0 generated`で3 groupすべてを再利用しました。Q3とQ4の復号後RGBおよびPCM SHA-256は一致しています。9枚Reference Imageの別Gateは下記に記録しています。これは実行、prefix再利用、AV再構築、テストした9枚経路の確認であり、画像・音声の総合的な主観品質評価ではありません。 + +### 9枚Reference Image入力Gate(0.3 MP参照画像) + +テストしたNVIDIA GeForce RTX 5060 Ti 16 GB/RAM 64 GB環境では、Reference Image 1~9、`512 × 608`出力、`3 × 5秒` Full Run、Spectrum Off、LoRA Off、Sage Attention、Balanced 22-frame continuity、各約`0.30 MP`のテスト参照画像で完走しました。Sampling中のVRAM使用量は、利用可能な`16,311 MiB`のうち約`15.0~15.5 GiB`であり、16 GB GPUでは余裕の小さい高負荷構成です。 + +9枚bundleは、追加の外観・ポーズ情報が必要で、十分なVRAM余裕を確保できる環境で使用してください。このsingle-seed Gateは実行と基本的な視覚的連続性を確認したものです。9枚が3枚より常に高画質であること、高解像度参照・大きな出力・別model・同時GPU負荷でも動作することを保証するものではありません。 + +16 GB GPUで9枚を使う場合は、読み込む前に元画像のコピーを約0.30 MPへ縮小してください。内部のconditioning用Resizeは継続しますが、9枚の大きな元画像によるメモリ負荷を管理する代わりにはなりません。 + +検証結果はテストしたローカルsourceと環境に限られます。導入済み環境が最新版であること、すべてのmodel/wrapperが任意のGPUに収まること、別のpromptでも同じ画質になることを保証するものではありません。 + +```text +python -m compileall -q . +pytest -q +``` + +MIT License。 diff --git a/docs/REGENERATE_INTENT_V396.md b/docs/REGENERATE_INTENT_V396.md new file mode 100644 index 0000000..fd21857 --- /dev/null +++ b/docs/REGENERATE_INTENT_V396.md @@ -0,0 +1,37 @@ +# Regeneration selections and automatic variations — 3.9.6 + +## Operation and compatibility + +V3.8 and V3.9 retain explicit Regenerate From and Variation Nonce across workflow save/reload and Queue submission. The intent marker is stored in workflow properties; public inputs, widget order and node IDs are unchanged. Choosing Auto clears the marker. The existing one-shot restart remains one-shot. + +With progress saving enabled, choose Full Run and an explicit Chunk. Nonce 0 lets Run Storage select an automatic variation: an interrupted compatible variation may continue, and queuing again after completion creates another variation. Fixed positive nonces and ordinary Full Run retain their existing deterministic identity and Core caching. + +Restart ComfyUI and refresh the browser after updating. Older unmarked Chunk 1 / nonce 0 values cannot be distinguished from old restart residue; legacy cleanup still restores Auto. To retain a deliberate Chunk 1 / nonce 0, select it again in the updated UI and save once. Old unmarked later-chunk or fixed-positive-nonce selections are retained. Do not delete Runs/Takes to update. + +Sampling v6, Graph v4, Run Storage v3 and State/Session formats are unchanged. This patch does not force resume_safe, relax reuse checks, rewrite unsafe/v5 history or migrate workflows. Run Storage decides whether a prefix is compatible. Earlier accepted chunks and previous Takes remain protected. All 25 distributed JSON/ZIP assets retain their bytes. + +## Verified scope + +Full real-Core CPU gate: 1,908 passed / 2 skipped / 2 existing Inspector expected failures / zero unexpected failures; CUDA initialization attempts 0. Production frontend fixtures cover 84 scenarios including 22 new restoration cases. Real-Core CPU cache checks cover ordinary Full Run, automatic nonce 0 and fixed positive nonces. + +Actual normal ComfyUI (Core 0.40.0, frontend 1.55.17): save a separate test workflow, reload the browser, allow real Run history to load, and verify Chunk 2 / nonce 5 and Chunk 1 / nonce 0 remain selected. The saved JSON includes the matching intent marker. This browser test is distinct from actual GPU prompts submitted through the normal Core API. + +| New-Run GPU Queue | Result | +| --- | --- | +| Review / Auto / nonce 0 | Generate Chunk 1 | +| Separate Queue, same Review input | Reuse Chunk 1, generate Chunk 2 | +| Full Run / Chunk 2 / nonce 5 | Keep Chunk 1, regenerate Chunk 2 | +| Full Run / Chunk 2 / nonce 0 | Keep Chunk 1, create automatic variation 6 | +| Identical Full Run / Chunk 2 / nonce 0 again | Keep Chunk 1, create automatic variation 7 | + +GPU: RTX 5060 Ti, PyTorch attention, Spectrum OFF, 512x512, Euler/simple, 8-step Turbo, two 5-second chunks, no First/Last/Reference inputs. Earlier Chunk 1 SHA-256 stayed identical; every newly generated Chunk 2 differed. The last two prompts, including output prefix, were exactly identical and the Sampler was absent from execution_cached. All manifests recorded Sampling v6 and resume_safe=true. All five video/audio outputs (first: 5 seconds; others: 10 seconds) fully decoded at 24fps. + +Existing saved workflows and reachable archived Run/Take files retained their recorded hashes. Older output paths unavailable before this repair were not counted as reverified; no missing-data diagnosis is inferred. Other attention/Spectrum combinations were not newly GPU-tested for this patch. Subjective audiovisual quality, mobile performance, full default-workflow acceptance and the known Inspector reconfigure issue remain separate. Previous Issue #31 coverage: ISSUE31_SAMPLING_SIGNATURES.md. + +## 日本語 + +明示したRegenerate FromとVariation Nonceを保存・再読込後も保持します。保存ON/Full Run/明示Chunk/Nonce 0では、毎QueueでRun Storageが継続か新Variationかを判断します。完了後の同一入力でも再生成し、前のChunkと既存Takeは保持します。固定の正Nonceと通常Full Runのキャッシュは維持します。 + +ComfyUI再起動とブラウザー再読込が必要です。旧markerなしChunk 1/Nonce 0は、更新後にChunk 1を選び直して一度保存してください。旧unsafe履歴は自動修復しません。既存WorkflowやRun/Takeの削除は不要です。 + +実画面でChunk 2/5とChunk 1/0の保存再読込保持、実GPUでChunk 1→2継続・Chunk 2再生成・同一Full Run入力2回の自動Nonce 6→7、元Chunk 1保持、動画5本全デコードを確認しました。今回のGPU範囲はPyTorch/Spectrum OFFです。Sampling v6の安全判定、既存配布Workflow、旧Takeを保護しています。 diff --git a/docs/RELEASE_NOTES_V3910.md b/docs/RELEASE_NOTES_V3910.md new file mode 100644 index 0000000..d7a90a2 --- /dev/null +++ b/docs/RELEASE_NOTES_V3910.md @@ -0,0 +1,67 @@ +# 3.9.10 — Spectrum reuse and Reference assignment display + +Fix two causes of false **Sampling v6 read-only** in the supported Spectrum +configurations: Windows CRLF source line endings and Python 3.12 implementation +signatures. Also correct the Reference Images assignment table when Chunks comes +from an external INT node. + +## Changes + +- Audit Spectrum CRLF and LF source line endings identically, including module + closures. All other source bytes and the exact supported file sets stay checked. +- Add separate audited Python 3.12 profiles for upstream Spectrum 0.2.29 at + `806fe498d574a726fecb8818917f00a4b993fd77` and the existing audited local + implementation. Retain the original Python 3.13 profiles byte-for-byte. +- With external Chunks inputs, display **assignment range 1–16** instead of a + stale Sampler widget count. All/Off actions identify their range. This range + does not set generation length or automatically expand saved selections. +- Preserve public node IDs, input/widget order, Sampling v6, Graph v4, existing + Takes and official workflow files. No chunk-LoRA or Project Manager feature is + included. + +## Update + +Update Continuum, restart ComfyUI and refresh the browser. Existing workflows need +no rewiring. Keep saved Runs/Takes. For a newly supported configuration, start a +**new Run**: older unsafe/v5 Takes remain readable and are not automatically made +reusable. Normal supported Python 3.13 generation identities stay unchanged. + +## Verification and limits + +- Full local CPU/JS gate: **2,096 passed / 2 skipped / 2 known Inspector expected + failures**; compilation and offline registration of all twelve nodes passed. +- Targeted Spectrum/Attention: **169 passed / 1 skip**. Python 3.12 component: + **112 passed / 1 skip**, with 21 independent older-Core/Attention cases excluded. +- Separate-process storage: **24 cases**, comprising 16 PyTorch/Kitchen cases on + Python 3.13 and eight Spectrum/default-model cases on Python 3.12. All matched + the actual Take read-only JavaScript. Original raw data was retained, and all + six previously valid full Sampling Contracts stayed identical. +- Reference table browser checks passed external INT and KJ Set/Get routing, + selection controls, disconnect/restore, save/reload and workflow tab switching. +- **No new GPU acceptance is claimed for 3.9.10.** These are CPU and UI checks, + not Attention-kernel, H3 inference, image-quality or performance measurements. + +Unknown source, changed functions, foreign closures and additional runtime state +remain non-reusable. The repair does not guarantee every Core/Kitchen/KJ/Sage or +Python combination. An independent Attention-observation failure in an older +Core/Kitchen environment can still disable reuse even when Spectrum is recognized. +The reporting user's exact environment has not been reproduced. + +See the [compatibility guide](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/blob/v3.9.10/docs/ISSUE31_SAMPLING_SIGNATURES.md) for audit details. + +## 日本語 + +対応済みSpectrum構成で、WindowsのCRLF改行とPython 3.12の実装識別によって +新規Takeまで「Sampling v6 read-only」になる判定を修正しました。外部INTでChunksを +指定したときのReference Images割り当て表も修正し、範囲1–16で設定できるようにしました。 +生成チャンク数や既存の選択内容を自動変更する機能ではありません。 + +更新後はComfyUIを再起動し、ブラウザーを更新してください。Workflowの配線変更や +履歴削除は不要です。新しく対応した構成は新規Runで確認してください。 +既存3.13の正常な識別値と、旧unsafe/v5 Takeの閲覧を維持します。 + +CPU/JS全回帰2,096成功、Python 3.12対象112成功、独立した保存・継続24ケースと +実JS照合、Reference表示のブラウザー操作を確認しました。 +**今回の修正としてのGPU受入は未実施です。** 未知の実装や追加状態の拒否は維持し、 +すべてのCore/Kitchen/KJ/Sage構成への対応や、報告者本人の環境再現を保証しません。 +4.0はメジャーアップデート用として温存し、SemVer互換の3.9.10で公開します。 diff --git a/docs/RELEASE_NOTES_V395.md b/docs/RELEASE_NOTES_V395.md new file mode 100644 index 0000000..0040a0a --- /dev/null +++ b/docs/RELEASE_NOTES_V395.md @@ -0,0 +1,21 @@ +# v3.9.5 — Issue #31 Attention and Spectrum Take reuse + +Fix finite identity observation of Core Model Attention Backend and slots-based SpectrumH3Binding so compatible new Runs can continue Review Each Chunk. Core PyTorch and Comfy Kitchen INT8 overrides include loaded implementations, closures and both container paths. Spectrum includes every effective config value and an explicit loaded-implementation contract; transient history/counters do not change identity. + +Includes a dedicated profile for upstream Spectrum v0.2.29 at 806fe498d574a726fecb8818917f00a4b993fd77, plus the prior audited implementation. Other source/runtime combinations remain non-reusable when their identity is unknown. Restart ComfyUI and start a new Run after updating; previously unsafe/v5 history is preserved and is not promoted or repaired. + +Validation: full local CPU gate 1,893 passed / 2 skipped / 2 existing Inspector expected failures / 0 failures; CUDA initialization attempts 0. Targeted signature gate: 152 passed / 1 scope-specific skip. Upstream 0.2.29 PyTorch+Spectrum and Kitchen+Spectrum both persisted Chunk 1 and reused it from a separate Queue to generate Chunk 2, with matching revision/contract, unchanged prefix hashes and fully decoded 512x512, 24fps, 10-second audio/video output. Actual CUDA backend calls were observed in the isolated gate. Prior normal-user-runtime checks also passed for the original audited Spectrum implementation. + +All 25 existing JSON/ZIP assets and 259 pre-existing Run/Take files were preserved. Sampling v6, Graph v4, Run Storage v3, twelve public IDs and official workflow bytes remain unchanged. This is targeted functional acceptance; browser interaction, subjective audiovisual continuity, BSA/Flow/Untwist, KJ/Sage patch-node combinations and benchmarks remain separate. See docs/ISSUE31_SAMPLING_SIGNATURES.md for environment-specific support. + +GitHub and Registry archives are checked for exact file sets, CRC and complete Manifest hashes. Hosted main and tag tests must pass before publication. Registry scan approval and real Manager installation remain separate from upload/archive verification. Prior tags, Releases and assets are retained. + +## 日本語 + +Model Attention BackendとSpectrumの署名を正確に識別し、新しいRunのReview Each Chunk継続を修正します。公式Spectrum v0.2.29も専用の監査済みプロファイルで対応します。更新・再起動後は新しいRunを作成してください。以前のunsafe/v5履歴は保持し、自動修復・再利用可能への昇格は行いません。 + +全CPU回帰は1,893成功・2スキップ・既知Inspector問題の期待失敗2件、失敗0・CUDA初期化0。公開Spectrum v0.2.29のPyTorch/Comfy Kitchen両条件で、別QueueからChunk 1を再利用してChunk 2へ継続しました。署名・元Takeのハッシュ保持・完成10秒動画の全デコードと実CUDA呼出しを確認しています。 + +既存25資産と259件のRun/Takeファイル、公開12ノード、Workflow、Sampling v6を保護しました。未知のSpectrumや異なる実行環境は安全側で再利用を拒否します。実画面・目視画質/音声・BSAやKJの組合せ全般の合格を主張するものではありません。 + +GitHub本体ZIP・チェックサム、Registry配布ZIPの整合性は公開工程で別々に確認します。Registry側の検査承認と実Manager導入は別判定です。3.9.4/3.9.3/3.9.2などの元タグ・Release・配布物は保持します。 diff --git a/docs/RELEASE_NOTES_V396.md b/docs/RELEASE_NOTES_V396.md new file mode 100644 index 0000000..38e0df1 --- /dev/null +++ b/docs/RELEASE_NOTES_V396.md @@ -0,0 +1,21 @@ +# v3.9.6 — Regeneration intent and automatic variations + +Fix Regenerate From / Variation Nonce restoration in V3.8 and V3.9. Explicit Full Run regeneration with progress saving and nonce 0 now reaches Run Storage on every Queue, so repeating the same input after a completed variation creates another Take while preserving earlier accepted chunks. Ordinary Full Run and fixed positive nonces keep their caching behavior. + +Restart ComfyUI and refresh the browser after updating. For an older unmarked Chunk 1 / nonce 0, select Chunk 1 again and save once. Existing Runs/Takes and official workflows do not need deletion or migration. Sampling v6 safety and unsafe/v5 history guards are unchanged. + +Validation: 1,908 local CPU passed / 2 skipped / 2 known Inspector expected failures / zero unexpected failures and CUDA initialization attempts. Actual browser save/reload retained Chunk 2 / nonce 5 and Chunk 1 / nonce 0 after history loading. Five new-Run GPU queues passed Chunk 1 creation, separate-Queue Chunk 2 continuation, Chunk 2 regeneration and two identical Full Run / nonce 0 submissions (automatic variations 6 then 7). Earlier Chunk 1 bytes were unchanged; all five video/audio outputs fully decoded. + +GPU scope: PyTorch attention, Spectrum OFF, RTX 5060 Ti, 512x512, 8-step Turbo, two 5-second chunks. Browser acceptance and Core-API GPU submissions are distinct checks. Subjective quality, other accelerators and the known Inspector issue are not newly accepted. Existing saved workflows, reachable archived Takes and all 25 JSON/ZIP distribution assets were hash-protected. Previously unavailable old output paths are not claimed as reverified. Details: docs/REGENERATE_INTENT_V396.md. + +GitHub ZIP/checksums and Registry archive pass exact file-set, CRC and Manifest checks during publication. Registry upload, service scan approval and actual Manager installation are separate. Earlier Releases/tags/assets are retained. + +## 日本語 + +V3.8/V3.9の再生成指定・Nonceを保存再読込後も保持します。保存ON/Full Run/明示Chunk/Nonce 0の同一入力でもRun Storageへ到達し、完了後は新Variationを生成できます。前のChunkと既存Takeを保持し、通常Full Runと固定Nonceのキャッシュは維持します。 + +更新後はComfyUIを再起動し、ブラウザーを再読込してください。旧markerなしChunk 1/Nonce 0は選び直して一度保存してください。既存WorkflowやRun/Takeの削除・移行は不要です。Sampling v6の安全判定と旧unsafe履歴の扱いは変更しません。 + +CPU 1,908成功・2スキップ・既知Inspector期待失敗2、失敗0・CUDA初期化0。実画面でChunk 2/Nonce 5とChunk 1/Nonce 0の保存再読込保持を確認。PyTorch/Spectrum OFFの新RunでGPU 5 Queueが成功し、Chunk 1→2継続・Chunk 2再生成・同一入力の自動Nonce 6→7・元Chunk 1保持・動画5本全デコードを確認しました。画質評価や他アクセラレーターの新規合格は主張しません。 + +GitHub配布物整合性とRegistryアップロードは公開工程で確認します。Registry検査承認・実Manager導入は別判定です。以前の元タグ・Release・配布物を保持します。 diff --git a/docs/RELEASE_NOTES_V397.md b/docs/RELEASE_NOTES_V397.md new file mode 100644 index 0000000..af0d6e8 --- /dev/null +++ b/docs/RELEASE_NOTES_V397.md @@ -0,0 +1,30 @@ +# v3.9.7 — External-input regeneration cache fix + +Fix automatic Full Run variations when Chunks or Variation Nonce come from external nodes. Unresolved linked inputs no longer let Core skip an explicitly requested regeneration. Earlier accepted chunks, existing Takes, public inputs and official workflows are preserved. Update the custom nodes, restart ComfyUI and refresh the browser; no workflow changes or history deletion are required. + +CPU: 1,948 passed / 2 skipped / 2 known Inspector expected failures; zero unexpected failures and zero CUDA initialization attempts. Targeted Core-API GPU queues verified external Chunks=2 / Nonce=0, repeated identical automatic variations 1 then 2, and 1 reused / 1 generated chunk per variation. Browser save/reload and Run-button submissions were checked separately. GPU scope: PyTorch attention, Spectrum OFF, 512x512, 8-step Turbo, two 5-second chunks. This is a targeted cache repair, not general image-quality or all-accelerator acceptance. + +Known ordinary Full Run, storage OFF and directly specified fixed nonces retain stable caching. An unresolved externally connected nonce may reevaluate the Sampler even when fixed; Run Storage still decides safe prefix reuse. Sampling v6 / Graph v4 and existing unsafe/v5 history safeguards are unchanged. + +Existing workflow JSON/ZIP assets remain byte-for-byte unchanged. GitHub asset integrity and Registry upload are verified by the publishing workflow; Registry scan approval and Manager installation remain separate checks. + +## 日本語 + +ChunksやVariation Nonceを外部ノードから接続した際に、明示したFull Run再生成がCoreのキャッシュで省略される問題を修正しました。前の採用済みChunk、既存Take、公開入力、公式Workflowは維持します。更新後はComfyUIを再起動し、ブラウザーを再読込してください。Workflowの変更や履歴の削除は不要です。 + +CPU 1,948成功、2スキップ、既知の期待失敗2件。外部INT接続で同一入力から自動Variationが進み、Chunk 1を保持してChunk 2だけ再生成することをGPUで確認しました。ブラウザーの保存・再読込とRun操作も別途確認しています。画質・全アクセラレーター構成の新規合格は主張しません。外部接続の固定Nonceはキャッシュ判定時に値が未解決のためSamplerを再評価する場合がありますが、保存済みprefixの再利用は既存の安全判定に従います。 + + +## Post-release verification on the published v3.9.7 tag + +**Published v3.9.7 GPU recheck (2026-10-15): PASS.** After restarting with the published tag (`18df80d`), external Core INT nodes supplied Chunks=2 and Nonce=0. Two completely identical Chunk 2 regeneration prompts advanced automatic Nonce **1 → 2**; both reused Chunk 1 and generated only Chunk 2. Earlier Takes and raw latents were retained. All three outputs (baseline plus two variations) fully decoded with 240 frames and 10-second video/audio. Existing 60 workflows, one settings file, 30 Run/Take files and all 459 release files remained hash-identical. Scope: RTX 5060 Ti 16 GB, 512×512, 24 fps, 2×5 seconds, 8-step Turbo, PyTorch attention, Spectrum OFF, Core API submissions. No CUDA/OOM/execution errors were observed; this does not extend browser, subjective-quality or other-accelerator acceptance. + +| Queue | Reused chunks | Generated chunks | Automatic Nonce | Elapsed | +| --- | ---: | ---: | ---: | ---: | +| Baseline | 0 | 2 | 0 | 177.8 s | +| Regenerate Chunk 2, first run | 1 | 1 | 1 | 90.4 s | +| Same input, second run | 1 | 1 | 2 | 90.5 s | + +### 公開版での追加GPU検証 + +**公開版3.9.7でのGPU再検証(2026-10-15):PASS。** 公開タグ(`18df80d`)で再起動後、外部Core INTからChunks=2・Nonce=0を入力し、完全に同一のプロンプトでChunk 2を2回再生成しました。自動Nonceは**1→2**へ進み、両回ともChunk 1を再利用してChunk 2だけを生成。以前のTake・raw latentも保持しました。基準生成を含む動画3本は240フレーム・映像/音声各10秒で全デコード成功。既存Workflow 60件・設定1件・Run/Take 30ファイル・公開版459ファイルのハッシュは不変です。条件はRTX 5060 Ti 16 GB、512×512、24fps、2×5秒、8-step Turbo、PyTorch attention/Spectrum OFF、Core API経由。CUDA/OOM/実行エラーは確認されず、ブラウザー・主観画質・他アクセラレーターの受入範囲は拡張していません。 diff --git a/docs/RELEASE_NOTES_V398.md b/docs/RELEASE_NOTES_V398.md new file mode 100644 index 0000000..9d97509 --- /dev/null +++ b/docs/RELEASE_NOTES_V398.md @@ -0,0 +1,47 @@ +# v3.9.8 — KJ Model Preview Override continuation + +Fix saved-Take continuation after an explicitly audited KJ Preview Override retains preview images. The dedicated adapter separates reviewed transient frames from fixed settings and implementation identity, preserving cross-key/list wrapper execution order. A separate Queue can reuse Chunk 1 and generate only Chunk 2. Sampling v6 safety, existing Takes, public inputs and official workflow bytes are preserved. + +**Limited Phase A support:** Preview Override VAE / Audio VAE disconnected, Tiny VAE `none`, Core NoPreviews/Auto/Latent2RGB, and only the exact provider/Core audit profile. KJNodes 1.5.2 alone is not sufficient; the audited local Preview Override source differs from upstream. Unknown or altered providers retain conservative non-reuse. See [the supported source hashes and update guide](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/blob/v3.9.8/docs/KJ_PREVIEW_OVERRIDE_V398.md). + +Restart ComfyUI, refresh the browser and start a new Run. Do not delete old workflows or Takes. Earlier safe KJ identities may differ; unsafe/v5 history is not repaired or promoted. Non-KJ v6 signatures are unchanged. + +CPU: 1,995 passed / 3 skipped / 2 existing Inspector expected failures; zero unexpected failures or CUDA initialization attempts. Dedicated real-provider tests: 48 passed, no skips. Four normal-runtime Core-API GPU Queues passed: PyTorch and Comfy Kitchen INT8 + audited local Spectrum, each saving Chunk 1 and continuing to Chunk 2 in a separate Queue with held previews. The KJ MODEL was cached on each second Queue, contracts stayed equal, and original Chunk 1 bytes were preserved. All four AV outputs fully decoded. + +GPU scope: RTX 5060 Ti 16 GB, 512x512, 24fps, 8-step Turbo, 2x5-second chunks, Core Auto preview, no preview VAE. The initial GPU gate used candidate product bytes; the subsequent published-tag recheck is recorded below. No new browser Run-button, per-kernel return trace, subjective-quality or performance acceptance is claimed. The Inspector issue remains. Source/tag CI and exact GitHub/Registry archive checks run in publication; Registry scan approval and real Manager installation are separate. + +## 日本語 + +監査済みKJ Model Preview Overrideで画像を保持した後、保存Takeを継続できない問題を局所修正しました。一時画像と固定設定を分離し、実装識別とラッパー順序、Sampling v6の安全判定を維持します。 + +対応は固定した監査プロファイルと、VAE/Audio VAE未接続・Tiny VAE none・Core NoPreviews/Auto/Latent2RGBに限定します。KJNodes 1.5.2全般を保証するものではなく、監査したローカルソースは公開upstreamと異なります。未対応構成は安全側で再利用不可を維持します。 + +更新後はComfyUIを再起動し、ブラウザーを再読込して新Runを作成してください。既存Take・Workflowは保持し、旧unsafe履歴を自動修復しません。CPU 1,995成功(専用48件を含む)、PyTorchおよびKitchen+ローカルSpectrumの4 Queueで別QueueのChunk 1→2継続、元Chunk 1保持、動画4本の全デコードを確認。画質・速度・全KJ構成の合格は主張しません。Registryの検査承認と実Manager導入は別確認です。 + +## Post-release verification on the published v3.9.8 tag + +### Published v3.9.8 GPU recheck — 2026-10-16: PASS + +After restarting the normal ComfyUI installation with the published tag **`v3.9.8` / `58fbd855`**, all 464 installed release files matched the tag. The startup log confirmed 3.9.8, and new saved Sampling v6 contracts contained the KJ adapter and pinned audit-profile digest. Two brand-new Runs passed four Core-API GPU Queues: + +| KJ Preview Override configuration | Separate-Queue Chunk 1 → 2 | +| --- | --- | +| Model Attention Backend: PyTorch, Spectrum OFF | **PASS — 1 reused / 1 generated** | +| Model Attention Backend: Comfy Kitchen INT8 + audited local Spectrum | **PASS — 1 reused / 1 generated** | + +On both second Queues, the KJ MODEL stayed cached with preview images retained (**8 → 16**), while the Sampler executed. The contract/revision and original Chunk 1 latent hashes stayed unchanged; `resume_safe=true`. All four video/audio outputs fully decoded (5 seconds / 120 frames, then 10 seconds / 240 frames, 512×512 at 24fps). Existing workflows and Takes were preserved: 170 protected files, including 71 saved workflow files and 46 pre-existing Run/Take files, retained their hashes. + +**Scope:** only the exact audited KJ/Core profile; the local KJ source differs from upstream, so KJNodes 1.5.2 alone is insufficient. Preview Override VAE / Audio VAE were disconnected, Tiny VAE was `none`, Core preview was Auto. GPU: RTX 5060 Ti 16 GB, 8-step Turbo, Euler/simple, 2×5-second chunks. Spectrum used the audited local offline-replay implementation, system-RAM history/archive and model-aware OFF. **Unverified:** browser Run-button/save-reload behavior, the reporter's exact complete graph, other KJ/Spectrum builds, SolAttn/preview VAE paths, per-kernel return tracing, subjective quality and speed/VRAM benchmarks. Retained PNGs were 32×32, not high-resolution preview-quality evidence. The known Inspector issue remains; old unsafe/v5 Takes are not repaired or promoted. [Issue #31 result](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/31#issuecomment-6209571261). + +### 公開版v3.9.8のGPU再検証 — 2026-10-16:PASS + +通常ComfyUIに公開タグ **`v3.9.8`/`58fbd855`** を反映して再起動し、公開版464ファイルとの一致、実ログの3.9.8読込、新しいSampling v6契約内のKJアダプターと固定監査プロファイルを確認しました。新Runを2つ使用した計4 Queueの結果です。 + +| KJ Preview Override併用構成 | 別QueueのChunk 1→2継続 | +| --- | --- | +| Model Attention Backend:PyTorch/Spectrum OFF | **PASS — 1再利用/1生成** | +| Model Attention Backend:Comfy Kitchen INT8+監査済みローカルSpectrum | **PASS — 1再利用/1生成** | + +両構成の2回目は画像を保持したKJ MODELをキャッシュ再利用(**8→16枚**)し、Samplerは実行。契約・revision・元Chunk 1 latentのSHA-256を維持し、`resume_safe=true`でした。映像・音声4本は全デコード成功(512×512/24fps、5秒120フレーム→10秒240フレーム)。保護対象170ファイル(保存Workflow 71件・既存Run/Take 46ファイルを含む)は全ハッシュ不変です。 + +**対応範囲:** 固定したKJ/Core監査プロファイルのみ。ローカルKJソースはupstreamと異なり、KJNodes 1.5.2全般の合格ではありません。Preview OverrideのVAE/Audio VAE未接続・Tiny VAE `none`・Core Auto、RTX 5060 Ti 16 GB、8-step Turbo/Euler/simple、2×5秒。Spectrumは監査済みローカル版のoffline replay、履歴・archiveはsystem RAM、model-aware OFFです。**未確認:** ブラウザーRunボタン/保存再読込、報告者の完全一致graph、別KJ/Spectrum版、SolAttn/preview VAE、カーネル単位の戻り値、主観画質・音質、速度・VRAM比較。保持PNGは32×32で、高解像度プレビュー品質の証明ではありません。既知Inspector問題は残り、旧unsafe/v5 Takeを自動修復・昇格しません。[Issue #31への結果返信](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/issues/31#issuecomment-6209571261)。 diff --git a/docs/RELEASE_NOTES_V399.md b/docs/RELEASE_NOTES_V399.md new file mode 100644 index 0000000..f586b50 --- /dev/null +++ b/docs/RELEASE_NOTES_V399.md @@ -0,0 +1,69 @@ +# 3.9.9 — Public KJ Preview Override and Review continuation + +Some new Runs were marked **Sampling v6 read-only** when using the upstream KJ +Model Preview Override implementation, even though generation completed. This +release adds a separately audited upstream profile and explains saved reuse +failures in Review and Render History. The original supported local profile and +unaffected generation identities are retained; the safety checks are not relaxed. + +## Update + +Update Continuum, restart ComfyUI and hard-refresh the browser. Existing workflows +need no rewiring. Keep all saved Runs/Takes. After correcting an unsupported +configuration, start a **new Run**: old unsafe/v5 Takes remain readable and are +not automatically promoted to reusable history. + +Supported upstream KJ commit: `17195fcfa8a05f601b766a0f09c9733284f47416`. +On **Model Preview Override**, leave its optional VAE and Audio VAE disconnected, +set Tiny VAE to `none`, and use Core NoPreviews, Auto or Latent2RGB. Keep the +normal generation VAEs connected. This is not blanket support for every KJ/Core +version or unknown wrapper. See [the exact profiles and recovery steps](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/blob/v3.9.9/docs/KJ_PREVIEW_OVERRIDE_V398.md). + +## Validation + +- CPU: **2,075 passed / 2 skipped / 2 known Inspector expected failures**, zero + unexpected failures. The 90 production-JS cases and eight actual Run Storage + vectors passed; these overlap the pytest total. Offline twelve-node, + PackedLayout and Fixed 3x5s checks passed with zero recorded CPU CUDA + initialization attempts. +- Upstream KJ GPU: two separate Queues, **512x512 / 24fps / 2 x 5 seconds**, + FL2VA INT8 convrot, Qwen3VL NVFP4 AWQ, 8-step Turbo/Euler/simple and explicit + PyTorch attention. Chunk 1 paused for Review; the next Queue reused it and + generated only Chunk 2. The same cached KJ wrapper held eight real preview + images before Continue, while MODEL and sampling signatures remained equal. +- Browser on normal ComfyUI: with the pre-existing audited local KJ profile, + restart and frontend reload succeeded. **Use This Take** and **Continue From + Here** were visibly enabled. Continue From Here followed by the browser Run + button completed Chunk 2: **one reused / one generated**. Core execution times + were approximately 100 and 92 seconds, including decode/save; they are not a + speed comparison. Opening Render History involved one manual click; the + subsequent Continue selection and Run submission were automated browser actions. +- Both GPU pairs preserved the original Chunk 1 raw SHA-256, original Take and + correct parent link. All four 5s/10s video/audio outputs fully decoded. The + upstream GPU pair also passed saved-latent finite-value checks. No execution + errors or OOM occurred. The observed upstream system VRAM maximum was 15,446 + MiB at roughly six-second intervals, not an instantaneous per-process peak. +- All 21 existing workflow JSON/ZIP files remain byte-identical. Existing user + Take/workflow files were preserved; there is no storage or schema migration. + +These are targeted continuation/Review checks on Core 0.42.1, frontend 1.55.18, +Python 3.13.12, torch 2.14.1+cu130 and RTX 5060 Ti 16GB. Full CPU validation also +uses real Core/provider imports. They do not establish visual/audio quality, +long-video quality, all accelerators, all KJ versions, or save/reload acceptance +of every shipped workflow. The separate Turbo Sample already present on main is +included unchanged; its six-step/4-step-LoRA pairing has no new GPU acceptance. + +## 日本語 + +公開KJ Preview Overrideを接続した新規Runまで「v6 read-only」になる対応漏れを修正し、 +再利用不可のTakeには保存時の理由と対処法を表示します。安全判定は緩めていません。 + +更新後はComfyUIを再起動し、ブラウザーを強制再読込してください。既存Workflowの配線変更や +Run/Take削除は不要です。旧unsafe/v5 Takeは保持されますが、自動的に再利用可能には +なりません。設定修正後は新Runで開始してください。Preview Override側の任意VAE/Audio VAEは +未接続、Tiny VAEはnoneが条件です。通常の生成用VAEは接続したまま使用します。 + +CPU 2,075成功、公開KJでの実GPU継続、既設KJを維持した実ブラウザーでのボタン有効化と +Continue From Here→Run→Chunk 2完了を確認しました。両GPU構成でChunk 1を保持し、 +1再利用・1新規生成、Takeの親子関係、映像・音声全デコードを確認しています。 +対象修正の検証結果であり、全KJ構成や画質・長時間品質への包括的な合格ではありません。 diff --git a/docs/REVIEW_RESTORATION_REPAIR.md b/docs/REVIEW_RESTORATION_REPAIR.md new file mode 100644 index 0000000..93f56b6 --- /dev/null +++ b/docs/REVIEW_RESTORATION_REPAIR.md @@ -0,0 +1,27 @@ +# Review restoration after workflow-tab switching + +## What changed + +Returning to a workflow could hide the Review actions or disable Take application while the saved Take data still existed. The frontend could record a settings snapshot before upstream nodes, links and deferred configuration had finished restoring, then mistake the completed restoration for a user edit. + +History loading now waits for graph restoration and the existing deferred configuration passes. Graph/node request guards discard callbacks and responses from obsolete or detached nodes. Initial history reads preserve saved Take-selection fields. Loading and failed-history states have distinct summaries. Transient Review widgets remain excluded from workflow serialization. + +Real generation-setting edits still hide the three Review actions and disable Use This Take / Continue From Here. Returning to the original settings restores these controls. No automatic Take application or generation is performed during restoration. + +## Verified locally + +- Real production JavaScript fixture: 60/60 cases (48 existing plus 12 lifecycle cases). The previous code fails seven cases in the expanded fixture. +- Focused CPU tests: 108 passed, no failures; one existing pynvml deprecation warning. This is not a new full CPU suite. +- Chrome: five V3.9-copy / V3.8X2-copy / V3.9-copy round trips, with the V3.9 Review actions restored on return to the V3.9 copy. Render History retained its one compatible Take. +- Chrome real-edit guard: change Base Seed by one; confirm the three Review actions disappear and Use This Take / Continue From Here are disabled. Restore the original seed; confirm all five controls become enabled again. +- Saved workflow and Run project.json hashes remained unchanged after the guard test; no Take was applied and no generation was queued. + +The browser fixture uses a historical Run whose stored generation settings differ from the loaded diagnostic workflow. It validates frontend restoration and edit guarding only, not backend Take compatibility or GPU continuation. Previous/Next remained disabled because the fixture had only one compatible Take. + +## Scope and update + +The runtime change is frontend-only in web/project_id.js. Sampling, Run Storage schemas, State/Session/Assembly contracts, official workflows, package version, historical Releases and Registry publication are unchanged. New regression tests and integrity manifests accompany the fix. + +Update the custom node from main and refresh the browser (hard-refresh if necessary) before testing. If the Review controls still disappear, report the ComfyUI/frontend versions and the exact tab-switch sequence, including whether settings were edited. Do not delete saved Runs or Takes to work around this UI issue. + +No GPU continuation or new GitHub Release / Registry publication was performed for this repair. diff --git a/docs/SEAM_TEST_SOURCE_JA.md b/docs/SEAM_TEST_SOURCE_JA.md new file mode 100644 index 0000000..2127a51 --- /dev/null +++ b/docs/SEAM_TEST_SOURCE_JA.md @@ -0,0 +1,56 @@ +# H3 Continuum Seam Test Source + +`H3 Continuum Seam Test Source (Experimental)`は、Video Seamの比較専用ノードです。 +H3 Samplingは実行せず、同じ入力画像から決定的な3つの5秒Chunkを作ります。 + +## 接続 + +```text +Load Image + image + -> H3 Continuum Seam Test Source (Experimental) + +Seam Test Source: images + -> H3 Continuum Assemble + Seam: images + +Seam Test Source: audio + -> H3 Continuum Assemble + Seam: audio + +Seam Test Source: assembly_plan + -> H3 Continuum Assemble + Seam: assembly_plan +``` + +Assembler側は次の設定にします。 + +```text +exact_total_duration = true +Audio Seam = Off +Report Detail = Detailed Report +``` + +## 比較手順 + +同じ入力画像と`Exposure Change = 0.08`を維持し、`Video Seam`だけ変更します。 + +1. `Analyze Only` +2. `Auto` +3. `Auto 2`(露出ランプ補正はExperimental) + +期待結果は次のとおりです。 + +```text +Analyze Only + -> 1->2、2->3とも class=exposure_ramp + -> フレームは未変更 + +Auto + -> exposure rampは補正対象外 + -> action=kept native boundary + +Auto 2 + -> 各境界の先頭4フレームだけ露出を補間 + -> action=normalized exposure/color on 4 exposure ramp boundary frame(s) +``` + +`Test Resolution = 256 px (Fast)`を推奨します。アルゴリズムの判定確認に高解像度は不要です。 +`Input Size`は800x800入力で数GiBのRAMを消費する可能性があります。 diff --git a/docs/TEMPORARY_STORAGE_JA.md b/docs/TEMPORARY_STORAGE_JA.md new file mode 100644 index 0000000..5b04d22 --- /dev/null +++ b/docs/TEMPORARY_STORAGE_JA.md @@ -0,0 +1,210 @@ +# H3 Continuum 一時ストレージ設定 — 実装仕様 v1 + +作成日:2026-10-13 + +基準:V3.9.2 / `e9b1294ccd08dfce972cd8e5a11703b3f616ee54` + +状態:V3.9.3の実装仕様。検証結果とプラットフォーム範囲は`../PACKAGE_VALIDATION.txt`を参照。 + +## 1. 目的と対象 + +Photoshopの仮想記憶ディスクに相当する、一時データの保存先指定を追加する。独自の仮想ディスクを作らず、ComfyUIサーバーに接続された別ストレージ上のフォルダーを利用する。 + +対象は次の二つ。 + +- FinalizeのDisk-backed IMAGEバッファ。 +- Decode Cache Helperのディスクキャッシュ。 + +Run Storage、採用済みchunk、Take、State/Session、完成動画・音声、他の拡張機能の一時ファイルは対象外。保存先変更で生成条件、解像度、dtype、Sampling/Reference/Reviewの契約を変更しない。 + +一時データの移動は書き込み先の変更であり、総書き込み量を減らす機能ではない。物理ストレージが1台しかない場合、別フォルダーや別パーティションへ移しても、その機器への書き込みを逃がせない。ただし十分なRAMがある場合のRAM利用や、任意のDecode Cache無効化は別の対策として利用できる。 + +## 2. 設定画面 + +ComfyUIの設定に「H3Continuum → Storage → 一時ストレージを管理」を追加する。専用ノードやワークフロー内のパスwidgetは追加しない。 + +| 項目 | 確定した動作 | +| --- | --- | +| 保存先フォルダー | サーバー側の絶対パスを入力。空欄で既存の保存先を継承 | +| ストレージ候補 | サーバーで認識できるローカルボリューム/マウント先と空き容量を提示。選択すると保存先入力を補助 | +| 検証・保存 | サーバー上でパス、書き込み、mmap、容量取得を検証してから保存 | +| 保存先と使用量 | Finalize、Decode Cacheごとに実際のパスと管理対象ファイル量を表示 | +| 空き容量 | 各保存先のボリュームの現在の空き容量を表示 | +| 使用中/削除可能/判定不能 | 一時ファイルの状態を分けて表示。判定不能なファイルは保護 | +| 旧保存先 | この設定で登録した旧保存先に残る管理対象ファイル量も表示 | +| 更新 | 手動更新。設定画面を開いた時と保存・削除完了時にも更新 | +| 未使用の一時ファイルを削除 | 実行中は無効。サーバー側でも再検査して、安全と確認できた対象だけ削除 | + +ブラウザー/モバイル端末のディスクではなく、ComfyUIが実行されているPC・サーバーのストレージを選ぶ。候補一覧が取得できないOSでも絶対パス入力で利用できる。SSD/HDDの種別を推測してラベル付けしない。異なるボリューム表示が異なる物理ディスクを意味するとは限らない。 + +表示するファイル量は論理ファイルサイズの合計。物理占有量、SSDの累計書き込み量、寿命指標ではないことを明記する。保存先の空き容量はファイルシステムから別途取得する。同じボリュームを使う二つの保存先の空き容量を加算しない。 + +描画callbackや常時タイマーでディスクを走査しない。画面表示・操作時に管理対象の範囲だけ確認し、ブラウザー側にイベントの重複登録を作らない。 + +API応答が404・空・不正な形式の場合は、その結果と再取得方法を表示する。候補欄は取得前から説明付きの幅を確保し、状態取得に成功するまで保存・削除を無効にする。更新で復帰でき、保存・削除リクエストを自動再送しない。起動ログにポート使用中エラーがある場合、旧プロセスが残って新しいPython側機能を起動できていない可能性を案内する。 + +## 3. 保存先の決定と既存動作 + +### 空欄 + +完全に既存の保存先を継承する。 + +- Finalize:ComfyUIのtemp配下の既存専用フォルダー。 +- Decode Cache:既存の`H3_DECODE_CACHE_TEMP_DIR`指定があれば継承し、なければOSのtemp。 + +空欄のまま、二つの保存先を勝手に統一しない。ComfyUI全体のtemp設定やOSの環境変数を書き換えない。 + +### 指定あり + +指定フォルダーの下に専用領域を作る。 + +```text +指定フォルダー/ + h3-continuum-scratch/ + finalize/ … 既存の.binと所有者情報を使用 + decode/ … helperごとの私有ディレクトリと.npy +``` + +新しい設定がDecode Cacheの旧環境変数より優先する。内部テスト等が明示的に渡す`backing_root`や`temp_parent`の依存注入は維持する。 + +v1の対応先はサーバー上のローカルファイルシステム。ネットワーク共有・UNC・リモートマウントは正式対応外とし、明確に識別できる対象は保存時に拒否する。同期ツールや特殊なファイルシステムの適合性は、書き込み・mmapテストだけでは保証しない。RAMディスクは通常のローカル保存先として扱えるが、専用の作成・管理機能は追加しない。 + +保存時に正規化したパス、ボリューム識別情報、専用領域の識別情報を記録する。各ファイル作成前にも確認し、マウント解除により元のパスが別ボリュームへ変わった場合は利用しない。 + +## 4. 設定の保存と切替 + +- サーバー共通の設定としてComfyUIのsystem user領域に保存する。ブラウザーのローカル保存だけでバックエンドを制御しない。 +- 設定本体と管理した保存先一覧は私有のJSONとし、atomic writeで更新する。インストール先、ワークフロー、公開READMEには利用者の実パスを保存しない。 +- パス変更と保存はQueueが実行中・待機中でない場合に限る。検証と保存を分け、検証だけで有効設定を変えない。 +- Finalize/Decodeの処理開始と設定切替・削除を共通の調停機構で制御する。各処理が取得した設定はその処理の終了まで保持する。 +- 保存完了後の新規ファイルから新しい保存先を使う。ComfyUI再起動を必須にしない。 +- 既存の一時ファイルは移動しない。旧出力のstorage/aliasやcache entryは、元の保存先と寿命管理を保持する。 +- Decode Cacheは保存先の切替後、新規entryを新保存先へ作る。切替時に既存の参照中entryを削除しない。旧entryの安全な再利用・退役を維持する。 +- 同じ設定ファイルを複数プロセスが変更する場合は、ファイルロックと設定revision照合で上書き競合を検出する。 +- v1ではマルチユーザー運用時のUIからの共有保存先変更・削除を無効にする。共有サーバーの保存先は管理者がサーバー設定として管理する。既存のユーザープロファイルを管理者認証とみなさない。 + +指定先が使えない場合、元のSSD・OS tempへ黙って戻さない。Finalizeでディスクが必要なら明確なエラーにする。Decode Cacheだけが利用できない場合は、既存の方針に合わせてキャッシュをバイパスしてnative Decodeを続け、元の保存先へキャッシュを書かない。中断・native Decodeの例外はバイパスで隠さない。 + +## 5. RAM/Diskの選択と事前確認 + +新しい設定画面に別のRAM/Diskモードを作らず、既存ノードの選択を維持する。 + +| 対象とモード | 確定した動作 | +| --- | --- | +| Finalize / Auto | 既存の4 GiB境界とRAM安全余裕を維持。ディスクを選んだ場合に新保存先を使用 | +| Finalize / RAM | RAMを使用。確保前に安全余裕を検査し、不足・取得不能なら明確なエラー。ディスクへ自動変更しない | +| Finalize / Disk-backed | 指定先を使用。確保前に空き容量を検査し、不足ならエラー。RAMへ自動変更しない | +| Decode Cache / Auto | 既存のディスクキャッシュ方針とquotaを維持し、ファイルの保存先だけ変更 | +| Decode Cache / RAM | 既存のRAM予算・headroom判定を維持。保持できないentryはキャッシュしない。ディスクへ逃がさない | +| Decode Cache / Off | キャッシュ書き込みを行わない。Finalizeの保存方式は独立 | + +Finalizeの必要バイト数`M`は、実際の出力shapeの要素数×実際のdtypeのバイト数で算出する。解像度と秒数だけの固定概算で判定しない。 + +RAM確保直前の判定は、既存のAutoと同じ余裕を基準にする。 + +```text +R = max(4 GiB, 物理RAM総量の10%) +利用可能RAM >= M + R +``` + +手動RAMにはAutoの4 GiB上限を適用しない。十分なRAMが確認できれば大きい出力も選択できる。Autoの4 GiB上限を引き上げる機能はv1に含めない。 + +Disk-backedの確保直前には、既存Autoと同じ`空き容量 >= M + 2 GiB`を確認する。設定保存時の小さなmmapテストは、将来の大容量確保が成功する保証として扱わない。 + +空き容量の検査は領域の予約ではない。検査後に別プロセスが容量を使う場合や、利用中のドライブが故障・切断する場合のI/O失敗を完全には防げない。メモリマップされた領域のI/O障害はOSによってプロセス終了につながる場合もある。v1で「誤設定やドライブ障害でもクラッシュしない」と保証しない。全領域を先にゼロ書き込みして二重の大量書き込みを発生させる対策は追加しない。 + +コンテナーのメモリ制限がある場合、ホストの空きRAMだけを根拠にしない。取得できる場合はプロセスに適用される制限と使用量も考慮し、より厳しい利用可能量を採用する。Autoの4 GiB/reserve方針自体は変えず、実効容量の取り方を検証する。 + +これはFinalizeの追加バッファを確保する前の確認であり、プロセス全体のOOM回避保証ではない。native VAE Decode、Sampling、下流ノードのコピー、並行処理、OSのページファイル利用まで制御しない。VRAMを一時ストレージにする機能は追加せず、GPUメモリ管理も変更しない。 + +RAM容量が取得不能な場合、Autoは既存どおりDisk-backedへ進む。手動RAMは安全確認不能として停止する。どちらの場合も選択理由と必要量を表示する。 + +## 6. 未使用ファイルの削除 + +「未使用」は生成が終わったことではなく、返したTensor/storage/aliasやcache mappingから参照されていないことを意味する。 + +- ブラウザーから削除パスを受け取らない。サーバーが登録・確認した保存先と、Continuum所有と確認できるファイルだけを対象にする。 +- 使用中のFinalize buffer、未公開の作成中buffer、参照中cache、別の生存プロセスのファイルを保護する。 +- Finalizeは既存のstorage寿命追跡、root lock、owner PID+process creation timeを再利用する。UI削除では、owner/sidecarが欠損・不正・判定不能なファイルを削除しない。 +- Decode Cacheはstore lockの下で未参照のディスクentryを管理表から退役させて削除する。単に.npyをunlinkして管理表を残さない。RAMキャッシュはこのボタンの対象外。 +- 参照中entryと削除保留entryはleaseが切れるまで残す。削除のために出力キャッシュを強制解放したり、gcを強制実行したりしない。 +- 新しいDecode Cache私有ディレクトリに所有者情報と排他制御を追加する。前回異常終了のファイルは、所有者が終了したと確認できた範囲だけ削除する。 +- 所有者情報がない既存のDecode Cacheディレクトリは、現在のstoreで管理しているファイル以外は自動削除しない。ファイル名のprefixだけを根拠にしない。 +- シンボリックリンク/junctionによる管理対象外への遷移をたどらない。設定先の解決結果を記録し、削除時に専用領域の同一性を確認する。 +- root・store lockと共通調停で作成/読み込み/削除の競合を防ぐ。他プロセスが現在書いている中間ファイルを削除しない。 +- 結果は削除数、削除した論理バイト数、使用中・判定不能・削除失敗の残数を報告する。失敗は保留し、使用中ファイルの強制削除へ切り替えない。 + +管理対象外のファイル、Run/Take、保存動画、他のComfyUI一時データは削除しない。一時領域全体の再帰削除を実装しない。キャッシュを削除すると次の利用時に再デコード・再書き込みが必要になり得ることも説明する。 + +## 7. 実装するファイルと責務 + +| ファイル | 具体的な変更 | +| --- | --- | +| `storage_settings.py` | サーバー共通設定、保存先の検証・解決、管理領域の識別、保存先一覧、調停、容量probeを実装。Tensorの強参照を持たない | +| `storage_api.py` | 設定表示・検証・保存・使用量・削除を提供。登録済み管理対象以外を削除しない | +| `web/storage_settings.js` | ComfyUI設定に操作画面を登録。保存先と実際の使用状態、エラー、削除結果を表示 | +| `storage_fs.py` / `storage_memory.py` | privateファイル・所有者・atomic write・プロセス間lock・cgroupを考慮したRAM取得を担当 | +| `__init__.py` | ComfyUI実ランタイム時だけAPIを一度登録。standalone/offline importを壊さない | +| `v3/file_backed_buffer.py` | 保存先resolverを共通設定へ接続。既存の明示`backing_root`、Auto方針と寿命管理を維持。手動RAM/Diskのpreflightを追加 | +| `v3/assembly_v35.py` | 新規出力確保前に保存先・メモリ・ディスクpreflightを呼ぶ。フレームコピー/Seam/音声の計算は変更しない | +| `tools/p20_file_backed_tensor_poc.py` | 読み取り専用の管理状況取得と、UI向け保守的削除経路を追加。alias寿命追跡と既存sidecarの互換性を維持 | +| `decode_cache/store.py` | 保存先切替、所有者記録、weakなstore登録、状態取得と未参照entry削除を実装。既存key・digest・dtype・quotaは維持 | +| `decode_cache/nodes.py` | helper実行開始時に有効な保存先設定を取得し、storeへ適用。入力widget・既定値・登録IDは変更しない | +| `README.md` / `README_JA.md` | 設定手順、1台構成の限界、HDDの速度、RAM・ページファイル、削除の意味を説明 | +| tests / manifest類 | 新しい安全性・互換性の試験を追加。実装時に既存配布manifestを更新・検証 | + +APIは`/h3-continuum/storage/`配下とし、`GET status`、`POST validate`、`POST settings`、`POST cleanup`で役割を分ける。UIは設定revisionを指定して保存し、古いrevisionや実行中の保存/削除は競合として拒否する。読み取りのstatusから削除やcache clearを起こさない。 + +既存のSampling・Run Storage・State/Session・Referenceの処理は変更しない。公式ワークフローJSON/ZIP、ノードID、socket/widget順序、既存のDecode Cache identityは維持する。Experimental 3、Review From Chunkの仕様は本件に含めない。 + +## 8. READMEへ必ず記載する内容 + +1. 別の物理ストレージへの移動は書き込み先を変える機能であり、総書き込み量を減らさない。 +2. 物理ストレージが1台の場合、保存先変更で別機器へ書き込みを移すことはできない。同じSSDの別パーティションも同様。 +3. HDDは使用できるが、Decode・結合・保存の速度が低下する場合がある。 +4. RAMは十分な余裕がある場合に限って利用する。事前確認があってもOOMやページファイルへの書き込みを完全には防げない。 +5. 一時保存先はComfyUIサーバー側。保存先を変えても旧ファイルは移動しない。 +6. 削除は容量の解放であり、既に発生したSSD書き込みを取り消すものではない。キャッシュ削除は次回の再書き込みを増やす場合がある。 +7. FinalizeとDecode Cacheは独立したモード。CacheをOffにするだけではFinalizeのDisk-backedを無効にしない。 +8. RAMディスクのデータは再起動で失われる。Run/Takeや完成動画を一時領域と混同しない。 +9. 現在の設定・直近の実行で選ばれたbackend・管理対象ファイル量を区別する。Autoを常にSSD使用と説明しない。 +10. RAM/ディスクの事前検査は安全性を高めるものであり、実行中のメモリ変動、ディスク容量変動・切断による失敗を完全には防がない。 + +実装前にスクリーンショットを作成しない。実装後、ユーザー提供の実画面をREADMEへ掲載する。 + +## 9. 検証と完了条件 + +### CPU/ファイルシステム + +- 空欄時のFinalize/Decode Cacheのパスと環境変数優先順位が従来と一致。 +- 指定時は両者の新規ファイルが専用領域に入り、旧保存先へ新規書き込みしない。 +- 日本語・空白を含むパス、Windowsドライブ、Linuxローカルマウントで動作。 +- 書き込み不可、mmap失敗、容量不足、指定ディスク切断、マウント先入れ替わりで明確な結果を返す。 +- 切替前の出力、Tensor view/alias、COW cache mappingが切替・削除後も読める。 +- 未参照ファイルだけを削除し、Run/Takeと対象外ファイルを保持。 +- owner不明、不正sidecar、symlink/junction、PID再利用、別の生存プロセスを保護。 +- Queue/Decode/Finalizeの開始と設定変更・削除の競合を実行順を制御して検証。 +- 異常終了後の新形式cacheの残骸は安全に回収でき、旧形式の未管理cacheは保護。 +- RAM不足・取得不能・コンテナー制限を小型Tensorとprobe差し替えで検証。実機OOMを起こして試験しない。 +- Autoの4 GiB境界とreserve、手動RAMの4 GiB超許可、手動Diskの容量検査を検証。 +- キャッシュ障害時のnative Decode継続と、Decode例外/中断の伝播を維持。 +- 小型入力で、保存先やbuffer backendを変えてもフレーム・音声・dtypeが一致。 + +### ブラウザーと互換性 + +- ComfyUI設定で保存・検証・更新・削除ができ、再接続/再読込後も保存先が一致。 +- 実行中・Queue待機中の保存と削除はUIだけでなくサーバーでも拒否。 +- 設定画面を繰り返し開閉してもlistenerやDOM、描画処理が増殖しない。 +- モバイルではサーバー側パスを扱うことが明確。 +- マルチユーザーでは設定・削除の変更操作を拒否。 +- 既存V3.8X2/V3.9ワークフローを変更せず保存・再読込できる。 +- 既存Review/Take/Reference、file-backed、Decode Cacheの回帰試験と標準validationが通過。 +- 公式ワークフローのbyte同一性、配布manifestの整合を確認。 + +CPU・ブラウザー検証で一時保存先機能の受け入れを判断する。実H3の長尺生成速度、映像・音声品質、実SSD書き込み量は別の実機評価であり、GPU未検証を隠して全体保証としない。 + +## 10. 初回実装へ含めないもの + +データの圧縮/半精度化、VRAMを一時ストレージとして使う機能、独自RAMディスク、複数ディスクへの自動分散・自動フォールバック、Auto RAM上限の変更、全フレームを持たないstreaming出力、ネットワーク共有の正式対応、実SSDの寿命推定は初回に含めない。 + +確定した初回実装は、保存先指定・状態表示・安全な未使用削除と、そのための容量/所有権/手動RAMの事前確認である。パッケージ番号は3.9.3。GitHub Release/tag・Registry公開はソース更新とは別の作業。 diff --git a/docs/TIMELINE_VIDEO_EXPERIMENTAL.md b/docs/TIMELINE_VIDEO_EXPERIMENTAL.md new file mode 100644 index 0000000..26668ff --- /dev/null +++ b/docs/TIMELINE_VIDEO_EXPERIMENTAL.md @@ -0,0 +1,83 @@ +# Unified Timeline Video — Experimental 1 + +## 状態 + +`main`向けExperimental機能です。実装基準ソース: `085943ad9cb8023a2042d4d731c83e638361c06e`。 + +ローカル受入はPASSです。Full CPU suiteは`1424 passed / 1 skipped / 0 failed`、実ComfyUIのブラウザ保存・再読込、Follow/RepeatのGPU生成、A/B mode分離、Manifest整合を確認しました。両GPU runは704×416、24fps、240フレーム、映像10秒、32kHz stereo音声10秒で完走し、OOM/NaN/allocation failure/crashはありません。Production標準への昇格やRelease/tag/Registry公開を意味せず、機能名どおりExperimentalです。 + +## 接続と設定 + +``` +Core Load Video + ↓ VIDEO +H3 Continuum Video Adapter (Force Rate = 24) + ↓ images / IMAGE +H3 Continuum Sampler V3.8 + └ Timeline Video Frames [内部名 reference_video_1 を維持] +``` + +Video AdapterとCore Load Videoは既存のままです。AudioはCore Load Audioを利用し、この機能が動画素材の音声を勝手に追加することはありません。 + +動画入力を有効にすると、Samplerに **Video Reference Mode** が表示されます。 +この設定は **Prompt Format = Timelineとは別**です。設定を増やすのはこの1項目のみで、新しい動画ソケット、Loader、Enableは追加しません。 + +### Follow Timeline + +生成全体の時間位置に対応する参照区間を渡します。20秒素材を2×10秒で使うと、前半グループは先頭側、後半グループはその続きです。 +厳密な境界はH3の17k+5フレーム整列と既存の出力計画に従います。例えば最初の10秒グループは243フレームとなり、次の参照開始は243/24=10.125秒です。単純な10.000秒への巻戻しではありません。 + +対象は **実際に新しく出力へ追加される自然長** です。Continuationの過去フレームprefixは二重に消費しません。 +Terminal Mergeでは、統合された1回のSamplingが担当する可視区間をまとめて使用し、Merge自体やSeedの決定を変更しません。 +Resume/Reviewでは保存済みのnet_framesを積み上げる既存カウンタを使い、Queueのたびに参照開始位置を0へ戻しません。 + +素材が終わったら、以後は動画参照なしで続行します。終端をまたぐ場合は残りを使い、残りが5フレーム未満ならそのグループの動画参照をスキップして診断を表示します。 +自動ループ、長い静止画保持、Stretchは行いません。通常のH3整列に必要な最大16フレームの短い終端補完だけは残します。 + +最後のグループを「今回要求した総尺」で参照クロップしない理由: 総Chunksを増やしたとき、保存済みグループの本来の参照入力が変わってしまうのを防ぐためです。実際の最終映像の尺調整は既存Assemblyのままです。 + +### Repeat Reference + +従来のVideo Guide Frames処理をそのまま使います。毎グループで同じ先頭側の、1チャンク上限内の素材を参照します。 +短い素材を必要尺まで何回もコピーする「ループ再生」ではありません。 +旧ワークフローや旧APIに新設定がない場合はこのモードになります。短い/長いという素材長だけでモードを自動変更しません。 + +### できないこと/保証しないこと + +- 元動画の動き、画素、人物、カメラを完全コピーする制御ではありません。参照映像は外観にも影響します。 +- 正しい区間を渡すことと、生成結果が意図した動きに一致することは別の評価項目です。 +- 単独15秒生成がstaticになるという報告やIssue #13の長尺driftの修正ではありません。 +- 長い入力のRAM削減機能ではありません。Followは全24fpsフレームの正規化済みCPUコピーを保持します。 +- 真のVFRタイムスタンプ追従や補間は追加していません。入力IMAGE列は24fpsという既存契約です。 + +## 内部構成 + +- `video_reference_modes.py`: Follow source、物理可視区間選択、Qwen/VAE共通の既存エンコーダ呼出し、slice identity。 +- `v3/driving_nodes.py`: V38へ省略可能なscalar widgetを末尾追加。V34入力境界でモードを消費し、Repeatは元の関数と元の引数を保ちます。 +- `v2/sequence.py`: 新sourceは既存の`reference_video_source`引数で運びます。旧`timeline_video_source`は使わず、Terminal Merge無効化を回避します。 +- 物理グループごとに別のprompt cache辞書を作り、同じ文章でも前の区間のconditioningを再利用しません。 +- VAE cacheの`source.combined_hash`にslice identityを入れます。共通キャッシュの実装は変更しません。 +- Run identityはFollow/Repeat、素材内容、サイズ、chunk_seconds、前処理・終端方針を区別します。Chunks総数は含めません。 +- 新規Followグループのplanに`_h3_continuum_video_reference_v1`を追加し、実参照範囲を記録します。Terminalの論理2区間には共通の物理slice情報が付きます。 +- UIは既存のネイティブwidgetを保存し、独自配列入替えを追加しません。古いワークフローの読み込み時は新モードをRepeatへ復元します。 + +## 配布ワークフロー + +適用ツールは元の公式JSONを変更せず、次の比較用コピーを生成します。各JSONと同じ内容のZIPも作成します。 + +- `MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Follow.json` +- `MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Repeat.json` + +変更はSamplerの新設定と動画入力の表示名のみです。元の画像、プロンプト、Seed、Run名、バイパス状態、外部ノード設定・接続は維持します。 +元の動画入力グループがBypassなら、そのままです。試験時にグループ全体を有効化してください。Core Load Videoだけを無効にしてAdapterを有効のままにしないでください。 + +## 受入結果と残る境界 + +1. Full CPU suite: `1424 passed / 1 skipped / 0 failed`。 +2. ブラウザ: Followを保存し、再読込後も`Video Reference Mode`と`Timeline Video Frames`を保持。 +3. Follow GPU: Chunk 1とChunk 2で別の参照区間を使用し、区間ごとに別conditioning/VAE cache identityを確認。 +4. Repeat GPU: 2 physical groupsで従来のpersistent参照を維持し、VAE conditioningはrun内で1回。 +5. A/B: Chunk 1はほぼ一致し、Chunk 2から明確に分岐。2 modeが同じconditioning経路へ潰れていないことを確認。 +6. Integrity: 出力仕様、AV duration、Manifest/Registry Manifest、OOM/NaN/crashなしを確認。 + +未検証の全モデル・解像度・長尺・source exhaustion・Resume/Retry組合せを一括保証しません。参照区間が正しく選ばれることと、生成映像が意図したmotionを忠実に再現することは別の評価です。Production標準への昇格、Release/tag、Registry公開は別判断です。 diff --git a/docs/TURBO_SAMPLE.md b/docs/TURBO_SAMPLE.md new file mode 100644 index 0000000..50bf21c --- /dev/null +++ b/docs/TURBO_SAMPLE.md @@ -0,0 +1,44 @@ +# V3.9 Turbo Sample + +A separate, user-supplied example of Continuum prompt formatting, not a new model +or a verified quality preset. Available in Templates as +`MiniMax_H3_Continuum_V39_Turbo_Sample` (no parenthesized suffix). + +- Prompt Format: Auto; 2 chunks x 8 seconds; Full Run; Progress OFF. +- The multiline prompt demonstrates `[Chunk 1]`, `[Chunk 2]`, first-image alignment, + continuous dancing with a fixed background/camera, and an off-screen male Thai rap. + `[Thai] ... ` is H3 lyric syntax, not part of the template name. +- Saved model: `minimax/DasiwaMinimaxH3_dasiwaHybridV4_int8.safetensors`. + Obtain it from the [author's model page](https://civitai.com/models/2877206?modelVersionId=3391862) + or select a compatible installed H3 model. No model weights are bundled. +- One LoRA is enabled: `minimax/minimax_h3_fl2v_turbo_4step_v1.2_768p_comfyui_bf16.safetensors`, + strength 1, from [LightX2V](https://huggingface.co/lightx2v/Minimax-h3-Turbo). + The supplied graph uses Euler / simple / 6 steps / Draft 0.30 MP. These are + preserved sample values, not a validated 4-step/768p configuration; check the + selected model and LoRA instructions before generation. +- Video VAE: `minimax_h3_video_vae_int8_convrot.safetensors`; audio VAE: + `minimax_h3_audio_vae_fp32.safetensors`; text encoder: + `qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors`, type `minimax`. + See [model download links and folders](../README.md#required-models-and-extensions). +- Select your own First Image (`example.png` is only a placeholder). The disabled + optional audio/video inputs retain sample filenames; select real files before + enabling them. Reselect model/LoRA paths if you do not use a `minimax` subfolder. +- Requires KJNodes (SageAttention and Model Preview Override), rgthree, and Spectrum + node definitions. Spectrum is OFF. Preview Override has video/audio VAE connected; + this is outside the audited v3.9.8 Preview Phase A Take-reuse profile. +- Thai lyric accuracy, unchanged background, identity preservation and subjective + video/audio quality are not established by adding this sample. Thai is outside + H3's explicitly listed stable dialogue languages. No new GPU test is claimed. +- Existing standard workflows and saved Runs/Takes are unchanged. Outputs use the + separate `video/MiniMax_H3_Continuum_V39_Turbo_Sample` prefix. GitHub main includes + this sample; the already-published v3.9.8 release/Registry package is not replaced. + +## 日本語 + +標準Workflowとは別のTurboサンプルです。Auto・2チャンク×8秒で、改行した +映像・環境音・音楽の記述、入力画像の指定、固定背景、画面外の男性タイ語ラップを +示します。画像・モデル・LoRAは各自で導入して選択してください。 +4-step/768p LoRAに対して6 steps・Draft解像度という添付の設定を保持しています。 +推奨設定や画質の合格を意味しません。タイ語の発音、人物・背景保持も未保証です。 +KJ Preview OverrideのVAE接続はv3.9.8の限定Take再利用検証の範囲外です。 +今回はサンプル追加とCPU検証のみで、実環境反映・新規GPU試験・Release更新は行いません。 diff --git a/docs/USER_GUIDE.md b/docs/USER_GUIDE.md new file mode 100644 index 0000000..4d00142 --- /dev/null +++ b/docs/USER_GUIDE.md @@ -0,0 +1,705 @@ +# H3 Continuum — illustrated user guide + +[Back to the quick start](../README.md) · [V3.9 References](V39_REFERENCE_INLINE.md) · [History](README_HISTORY.md) + +This guide preserves the detailed operating instructions and screenshots from +the README at source commit `894869d9` (package 3.9.8). +V3.8-labeled controls and screenshots describe the retained V3.8X2 workflow; +the V3.9 setup and Reference sections identify the separate current workflow. +Earlier dated fixes, version statements and measurements are historical evidence, +not a new acceptance claim. Use the [current README](../README.md) and the +[v3.9.8 compatibility profile](KJ_PREVIEW_OVERRIDE_V398.md) for current scope. + +- [V3.9 template setup](#start-here-v39-and-the-retained-v38x2-workflow) +- [Reference Images](#v39-reference-images) +- [Complete V3.8 control reference](#complete-v38-ui-reference) +- [Review, retry, Takes and extension](#review-continue-and-revise-a-chunked-video) + + + +## Start here: V3.9 and the retained V3.8X2 workflow + +The ComfyUI **Templates → ComfyUI-H3-Continuum** gallery offers two separate entries: the user-selected [V3.9 workflow](../examples/workflows/MiniMax_H3_Continuum_V39.json) and [V3.8X2 workflow](../examples/workflows/MiniMax_H3_Continuum_V38X2.json). The [V3.8X2 workflow ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2.zip) contains **both unchanged JSON workflows** for one-download setup; the [V3.9 ZIP](../examples/workflows/MiniMax_H3_Continuum_V39.zip) contains only V3.9. These are workflow archives, not custom-node installers. The V3.8 Sampler remains available, so existing V3.8X2 workflows still load. V3.9 does not automatically convert their Reference connections or saved Runs/Takes: preserve the old workflow and open the dedicated V3.9 graph for new work. + +The selected V3.9 template has 32 nodes. Image loaders 1–6 are wired to the new helper and bypassed by default; slots 7–9 remain available but require the user to add and connect their own image loaders. Choose actual files and enable only the desired images. The saved size controls are **First Image + Draft — 0.30 MP**, and the separate First Image loader is enabled; select your own First Image file before Queue. The saved 480×640 Width/Height values apply only in Manual mode. `Reference Image Size` is `Match Output`; it resizes conditioning after source-image load, not the source files themselves. For a clear per-chunk Reference test, turn First Image and Last Image off, switch to Manual, and use Reference Images alone. First/Last additionally constrain the video; Last Image may merge the final two logical chunks into one physical group, which requires the same effective Reference set in both. A subject may remain visible in later chunks through video continuity even after its Reference is no longer assigned. + +In the helper, `All chunks` uses connected images throughout; `Per chunk` provides an image-by-chunk table. The template starts in `Per chunk` with all assignments off: after enabling a loader, tick the chunks that should use it. Slot numbers never shift: Image 1 is `@R1` and Image 9 is `@R9`, even when intervening slots are empty. In Sequence Prompt, describe the action for each chunk and use `@R` tags only for active References; their actual `` numbers are resolved per group. Unused tags warn without stopping generation. Start with short chunks and pre-resized images: source images may consume RAM before the Sampler's Reference Image Size setting applies. + +V3.9.3 is the current source version on `main`. Download versioned source from [GitHub Releases](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases); ComfyUI Registry publication is separate. Historical Releases and tags remain available. Configured 1024×1024 Reference-only GPU/API runs succeeded, but the unchanged default template and browser save/reload remain separate acceptance gates. For a reproducible older environment, use the matching historical Release/tag and workflow; the retained V3.8 Sampler on current `main` specifically supports V3.8X2 workflows. + +`pyproject.toml` declares ComfyUI `>=0.32.0`; earlier real-generation/GPU validation used ComfyUI `0.34.2`, and the scoped Fixed cache comparison above used `0.38.0`. The minimum declaration does not establish V3.9 runtime acceptance on 0.32.0. The previous Windows CPU run is `1696 passed / 1 skipped / 0 failed`; the earlier separate Linux rerun report records `1534 passed / 3 skipped / 0 failed` and does not include the new cache or external INT tests. Neither replaces browser save/reload or GPU acceptance for the unchanged official template. + +## V3.9 Reference Images + +### Reference Images node: choose the image for each chunk + +![H3 Continuum Reference Images V3.9 node with fixed image slots and a per-chunk assignment table](images/v39/v39-reference-images-node.png) + +*Example UI, not the downloadable workflow's default assignments.* Image 1–9 are fixed slots: Image 1 is always `@R1`, even when other slots are empty. With `Reference Use = Per chunk`, tick the image/chunk intersections you want; each row's `All` / `Off` buttons select or clear that image across chunks. In this screenshot, Image 1 is assigned to Chunks 1 and 2, while other images have different assignments. The official V3.9 workflow starts with its image loaders bypassed and every assignment unchecked; choose real image files, enable their loaders, then tick the intended chunks. + +### Sampler V3.9: connect the helper once + +![H3 Continuum Sampler V3.9 with one Reference Images Optional input](images/v39/v39-sampler-node.png) + +Connect the helper's green `reference_images` output to the Sampler's single `Reference Images (Optional)` input. The V3.8X2 workflow uses its own V3.8 Sampler and older Reference sockets; opening it does not convert those connections to V3.9. This screenshot shows a **Manual 480×640 example**, not the official V3.9 workflow's saved **First Image + Draft 0.30 MP** setting. `Reference Image Size = Match Output` controls Reference conditioning size after image loading; it does not pre-resize source files or set output resolution. + +V3.9 now connects all nine fixed Reference Images through one **H3 Continuum Reference Images V3.9** node. Its in-node table selects All chunks or specific chunks. The V3.9 Sampler has one typed Reference Images connection and keeps Sequence Prompt, First/Last, Review and the rest of its generation controls. Use `@R1` for Image 1; the backend converts active tags to the group's actual ``. Both modes use the same group-routing engine. + +Prompt parsing never blocks generation. If an explicit Timeline cannot be parsed, the original text falls back to Fixed with diagnostic `H3C-P100`. An uncovered Timeline chunk reuses the previous prompt and reports `H3C-P101`. Select the image for each chunk in the Reference Images node; `@R1`–`@R9` in Sequence Prompt refer to those fixed image numbers. + +For the supplied `2 × 10 s` workflow, put each time header on its own line and describe only that chunk's action. For example, after enabling the matching image loaders and checking their chunk assignments: + +```text +[0-10s] +@R1 walks forward through a quiet hallway. The camera follows smoothly. + +[10-20s] +@R2 enters from the right. The camera turns to follow @R2. +``` + +Use the actual subjects and actions from your images; this is a syntax example, not a saved prompt in the downloadable workflow. Do not place prompt text on the same line as `[0-10s]`: that is not a Timeline header. A Reference can still influence a chunk without an explicit `@R` mention when its checkbox is enabled. + +The [official V3.9 workflow](../examples/workflows/MiniMax_H3_Continuum_V39.json) is the latest user-supplied JSON, copied without editing; its [single-workflow ZIP](../examples/workflows/MiniMax_H3_Continuum_V39.zip) is also available. It preserves the full V38X2 model/LoRA, Decode Cache, Finalize and Save chain. Six image loaders are prewired (initially bypassed); to use Image 7–9, add `H3 Continuum Load Image` nodes and connect them to the matching helper slots. The Sequence Prompt input remains connected but is intentionally blank: write your own prompt before Queue. The saved graph uses `Timeline`, two 10-second chunks, `Full Run` (displayed as Generate Full Video), First Image + Draft 0.30 MP size selection, and a separate Project ID. The earlier [Reference Inline migration fixture](../examples/workflows/MiniMax_H3_Continuum_V39_Reference_Inline.json) remains available unchanged. See the [step-by-step Reference Images guide](V39_REFERENCE_INLINE.md). Replace local media and model filenames as needed before Queue; enabling an image requires choosing a real file. Seven to nine active references can be particularly memory-intensive: start with fewer, smaller source images. A full ComfyUI backend restart and browser hard-refresh are required before the new node definitions appear. Browser save/reload and GPU acceptance of this unchanged default graph are still pending; a GitHub Release does not imply these gates or a Registry package have been completed. + +**Compatibility:** Existing V3.8X2 workflows still use the V3.8 Sampler and old Reference Images 4–9 helper. V3.9 needs its new helper and connection. Open the new V3.9 workflow instead of only changing the Sampler type in an old graph: older V3.9 prototype workflows with direct 1–3 connections and Sampler-side routing widgets need manual rewiring, and their saved widget positions may differ. Accepted V3.8 Runs/Takes are not automatically migrated. Keep a copy of the previous workflow and use a new Run Name for V3.9. + +Earlier browser checks covered the previous V3.9 controls. The new helper has GPU/API execution evidence; its browser save/reload and the unchanged default template's full acceptance are still open. CPU/JS and manifest results are separate; no V3.9 Release/Registry publication is implied. + +**Version boundary:** V3.8X2 package 3.8.3 and V3.9.0 are separate workflows. This update does not create a V3.8X2 or V3.9 GitHub Release, tag, maintenance branch, or Registry version; it does not remove any older Release. The existing [`v3.8.0` tag](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/tree/v3.8.0) remains the reproducible V3.8.0 baseline. + +## Loader persistence repair on main (2026-09-21, after 3.8.3) + +The current paired V3.8X2 workflows now use **Core Load Audio** and **Core Load Video -> H3 Continuum Video Adapter**. The adapter has Force Rate (default 24) and IMAGE/AUDIO outputs, but **no Enable switch or file selector**. Core owns video selection/upload/save; the established duration-preserving frame selection is reused and audio is passed unchanged. To disable Video Guide, bypass the adapter or the whole video-input group; do not bypass only Load Video while leaving its required-input adapter active. + +**H3 Continuum Load Image remains**, including Enable Image. Its mode observer now delegates to inherited Core setters (Issue #23), and no longer replaces the widget collection while drawing or saving. Enable is appended after Core's widgets; Core handles the filename and native Bypass persistence. The older Audio/Video node IDs remain loadable as deprecated compatibility nodes, not as the recommended new-workflow loaders. The V3.8X2 baseline registered ten node IDs including these two legacy nodes; V3.9 adds the V3.9 Sampler and Reference Images helper as separate IDs, for twelve total. V3.8 Sampler inputs, generation parameters, Decode Cache, Finalize and existing latent/storage contracts are unchanged. + +After updating, restart ComfyUI and reload/refresh the frontend. A workflow that already omitted its filename or saved an incorrect mode cannot recover those missing values: reselect the intended file and Bypass state, then save once. Existing user workflows and saved Takes are not automatically rewritten. Changing upstream node types can legitimately create a new Run Storage revision; preserve the original workflow when resuming accepted runs. + +Validation details and limits: [Loader repair record](LOADER_PERSISTENCE_REPAIR.md). CPU/store-model regression tests are not a substitute for the pending Windows Core 0.36.0 / frontend 1.53.6 browser tab-switch acceptance. No Release/tag or Registry publication is implied by this main-branch repair. Older loader screenshots below describe the retained legacy nodes. + + +## Timeline Video — Experimental on main + +The existing video IMAGE input is now displayed as **`Timeline Video Frames`** and has one appended optional control, **`Video Reference Mode`**. Existing workflows and API calls that do not contain the new setting keep the old behavior through the default **`Repeat Reference`** mode. + +- **`Repeat Reference`** keeps the legacy Video Guide contract: each physical generation group receives the same bounded prefix from the source. It does not loop or stretch a short source. +- **`Follow Timeline`** uses the 24 fps source interval corresponding to each physical output group. Continuation-prefix frames are not consumed twice. If the source ends, the remaining groups continue without video reference; a final usable partial interval may receive only H3's short frame-grid padding. + +This setting is independent of `Prompt Format = Timeline`. It does not add a new socket or loader, does not use the source video's audio automatically, and does not provide frame-perfect motion copying. The complete decoded IMAGE batch is still held in system memory, so this feature is not a long-input RAM optimization. + +Use the existing connection: + +```text +Core Load Video -> H3 Continuum Video Adapter -> Timeline Video Frames +``` + +The two comparison workflows are intentionally separate from the unchanged official V3.8X2 workflows: + +- [Follow Timeline JSON](../examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Follow.json) | [ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Follow.zip) +- [Repeat Reference JSON](../examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Repeat.json) | [ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2_Timeline_Experimental_Repeat.zip) + +Local acceptance passed with full CPU `1424 passed / 1 skipped / 0 failed`, browser save/reload, and Follow/Repeat `2 x 5 s` GPU runs at 704x416, 24 fps, 240 frames, and 10-second 32 kHz stereo audio. Follow changed to the later source interval for Chunk 2 while Repeat retained the shared prefix; both completed without OOM, NaN, allocation failure, or crash. Peak observed VRAM was about 15.6–15.7 GiB on the tested RTX 5060 Ti 16 GB system, so 16 GB headroom was small. This is a functional mode-separation result, not a universal speed or subjective-quality guarantee. See [Timeline Video Experimental](TIMELINE_VIDEO_EXPERIMENTAL.md). + +> **V3.8X2** is the product and workflow label for package `3.8.3`. It keeps the V3.8 production sampler and adds optional Reference Images 4–9 plus the built-in Decode Cache Helper. The older V3.8.0 package remains available from tag [`v3.8.0`](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/tree/v3.8.0). + exec-9cfa73b4-c1d3-4416-8950-fa661b946638_2 + + +## Official V3.8X2 workflows + +- **Standard name:** [JSON](../examples/workflows/MiniMax_H3_Continuum_V38X2.json) | [ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2.zip) +- **Helper-explicit name:** [JSON](../examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.json) | [ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2+Decode_Cache_Helper.zip) + +The two V3.8X2 JSON names contain the **same official graph**; they are distribution aliases, not different configurations. Both include the built-in Decode Cache Helper with its saved `Auto / RAM 256MB / Disk 8GB / reset_token 0` settings and the nine-reference helper connection. The standard V3.8X2 ZIP now also contains the separate V3.9 JSON, while the helper-explicit ZIP contains only its correspondingly named V3.8X2 JSON. Older V3.8X workflows remain available for existing projects. + +## Nine Reference Images: memory warning + +Reference Images 1–3 connect directly to the Sampler. Optional images 4–9 connect through `H3 Continuum Reference Images` and then to `Reference Images (Optional)`. Using all nine is a high-resource configuration, not the normal minimum requirement. + +The acceptance run completed on an **RTX 5060 Ti 16 GB / 64 GB system** only after the nine source references were kept near **0.30 MP each**. With `512 × 608` output, `3 × 5 seconds`, Spectrum Off, LoRA Off, Sage Attention, and Balanced 22-frame continuity, Sampling used about **15.0–15.5 GiB of 16,311 MiB VRAM**. There was little 16 GB headroom. + +When using many references, especially all nine, resize copies of the source images to about 0.30 MP before loading or connecting them. The Sampler still performs its normal conditioning resize, but smaller source files reduce input and preprocessing pressure. Larger references, larger output sizes, extra wrappers, or concurrent GPU work may require a higher-spec GPU and more system RAM. Nine references are optional and are not guaranteed to improve quality over three. + +## Built-in Decode Cache Helper + +### Important: this is not an unconditional new-generation speedup + +**Manual cache clear:** normally no action is needed. Press the Helper's "キャッシュをクリア" (Clear cache) button to discard that Helper's cache on the next Queue and run native Decode again. It does not delete immediately or queue automatically. The internal `reset_token` remains serialized for workflow/API compatibility and increments once per click (wrapping to 0 at the INT limit). Creating, loading, or cloning a node does not change it. Without JavaScript, or in API workflows, change `reset_token` through the traditional numeric input. + +**The Helper saves time only when unchanged latents are decoded again. It does not accelerate Sampling.** A first MISS runs native Decode and stores the result; hashing and storage can make that run slower. Generating different latents with a new random seed each time normally produces MISS, so this feature should not be expected to speed up that usage. + +Concrete reuse scenarios: + +- **Continue by adding chunks:** when existing chunk latents are retained and decoded again for output, cached existing chunks can HIT; newly generated chunks initially MISS. +- **Fix / partial regeneration:** unchanged chunk latents can HIT when submitted again; regenerated chunks whose latent contents changed MISS. +- **Complete / Resume output:** decoding retained latents again without resampling can benefit. Merely playing an existing video file is not a cache use case. +- **Retry from Decode onward, or change Finalize/output settings:** reuse applies when the same latents and decode-related metadata reach the Helper and Decode runs again. +- **Regenerate with a fixed seed:** reuse requires identical resulting latent contents. The same seed alone does not guarantee HIT. + +The action names themselves do not enable reuse. **Each Video/Audio entry must match in latent contents, shape, dtype, decode-related metadata, and VAE identity, and its cache entry must still exist in that Helper.** If ComfyUI reuses an existing output without executing Decode, there is no additional Decode saving attributable to this Helper. Connect the Helper and select `Auto` or `RAM`. In its report, `hit` means reuse, `miss` means native Decode, and `off` / `bypass` mean no cache reuse. + +Cache reuse is process-local and does not survive Python restarts. Eviction and VAE identity changes can also cause MISS. Changing mode, RAM/Disk budgets, or `reset_token` clears this implementation's Helper cache. Keeping `reset_token = 0` does not reset it on every Queue. + +V3.8X2 ships `H3DecodeCacheHelper` inside this Continuum package. It remains a separate public node under `MiniMax H3/Continuum/Helpers`; no second custom-node addon is needed. The original nine IDs remain registered for compatibility; the loader migration adds H3ContinuumVideoAdapter. Sampler, Finalize, Assembly Plan, Core Decode, and saved Sampler widget/socket contracts are unchanged. + +Connect Sampler `video_latents` and `audio_latents` plus their native Video/Audio VAE inputs to the Helper, then connect Helper `images` and `audio` to Finalize. Keep the Sampler `assembly_plan` directly connected to Finalize. The [official V3.8X2 workflow](../examples/workflows/MiniMax_H3_Continuum_V38X2.json) preserves its saved graph and settings. It contains no Core VAE Decode nodes, so returning to Core direct Decode requires adding those two Core nodes and reconnecting Finalize, or loading a prior Core-direct V3.8X workflow. Do not simply remove the Helper and expect automatic rewiring. + +`Auto` caches full physical Video decodes on private process-local disk and small Audio decodes in bounded RAM. `Off` delegates native Decode and clears the Helper cache. Cache failure falls back to native Decode; native Decode, OOM and queue interruption are not retried. Files are not reused across Python restarts. Keep only one provider of the `H3DecodeCacheHelper` node ID: the previous standalone addon must be disabled or removed during the migration, after a verified built-in installation. + +At 736×416 (0.306MP), 3×5 seconds, Euler/simple 6 steps, Turbo FL2V v1.2, a warm backend and three Video HITs, one measured configuration produced a **mechanistic estimate of about 47 seconds / 24.5% less Full Run Total**. One direct A/B pair observed 50.887 seconds / 26.46%, and two Complete/Resume re-decode pairs averaged 46.309 seconds / 79.826% less Total. These are narrow observations, not repeated universal averages or an unconditional 20%+ speed promise. The first MISS can be slower, new Sampling is not accelerated, and a changed latent or VAE identity produces a MISS. + +**Historical V3.8X workflow: [JSON](../examples/workflows/MiniMax_H3_Continuum_V38x.json) | [ZIP](../examples/workflows/MiniMax_H3_Continuum_V38x.zip)** — [Published releases](https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum/releases) + +✅ V3.8X on `main` includes the September 8, 2026 Review hotfix: Review Each Chunk continuation, stale `Regenerate From` state, completed-sequence extension, and Render History queue handling are repaired. Update with `git pull --ff-only origin main` (or ComfyUI Manager **Update**), restart ComfyUI, and hard-refresh the browser if the old UI remains. Existing saved Takes are preserved. Issue #13 remains a separate open long-continuation quality issue and is not part of this hotfix. + + +H3 Continuum is a Production Sampler for generating, reviewing, partially regenerating, and resuming long-form MiniMax H3 video without restarting the entire work. V3.8 has two product layers: **Main / Production** and **Advanced**. + +## Install or update first + +For a new Git installation: + +```bash +cd ComfyUI/custom_nodes +git clone https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum.git +``` + +For an existing Git checkout: + +```bash +cd ComfyUI/custom_nodes/ComfyUI-H3-Continuum +git pull --ff-only origin main +``` + +Restart ComfyUI after cloning or pulling. If ComfyUI Manager installed the node, use its **Update** action instead of mixing Manager updates with a second Git checkout. + +To install the historical V3.8.0 package into a separate checkout instead: + +```bash +cd ComfyUI/custom_nodes +git clone --branch v3.8.0 --single-branch https://github.com/ukr8b3g-cmyk/ComfyUI-H3-Continuum.git ComfyUI-H3-Continuum-v3.8.0 +``` + +Use only one Continuum checkout at a time. ComfyUI Manager **Update** tracks `main`; it does not select historical tags. + +Download the current V3.8X2 workflow: [JSON](../examples/workflows/MiniMax_H3_Continuum_V38X2.json) or [combined V3.8X2/V3.9 ZIP](../examples/workflows/MiniMax_H3_Continuum_V38X2.zip). The V3.8X2 JSON is one Spectrum-capable graph, also usable with [LightX2V Turbo](https://github.com/ModelTC/Minimax-H3-Turbo); both Spectrum and all Turbo LoRA entries are saved disabled. Install its external Spectrum, rgthree, and KJNodes nodes before opening it. ComfyUI-Easy-Use is not required. See [Spectrum and Turbo setup](#turbo-lora-and-spectrum-in-supplied-workflows) below. + +## Prompt and skill downloads + +- [V3.9 Reference Images prompt skill](../write-h3-v39-reference-prompts/SKILL.md) | [ZIP](../write-h3-v39-reference-prompts.zip) — for fixed `@R1`–`@R9` image slots and per-chunk assignments. Use it with the V3.9 Reference Images node; it is separate from the general Continuum prompt skill below. +- [LLM system prompt ZIP](../H3-Continuum-LLM-System-Prompt-v1.zip) — system instructions and reference material for Continuum prompt authoring. +- [Continuum prompt skill ZIP](../H3-Continuum-Skill-v1.zip) — general chunk-aware prompt authoring for Codex and other compatible agents. +- [Continuum Dance Director skill ZIP](../minimax-h3-continuum-dance-director.zip) — long-form choreography prompts for one primary dancer, with body and camera continuity across chunks. + +These are optional prompt-authoring resources, not ComfyUI custom nodes. Extract each ZIP and follow its included instructions. They do not change the sampler or guarantee generated motion or image quality. + +## You can ask an AI to read this manual + +This README is intentionally detailed. You do not have to memorize it. Give its GitHub URL or this file to a web-enabled AI, describe your target video, GPU/VRAM, available First Image/audio, and whether you want to review every chunk. Ask it to use the **exact V3.8 labels** below. + +Example request: + +```text +Read the current ComfyUI-H3-Continuum V3.8 README. I have a 16 GB GPU and want +to create a 30-second I2VA video, reviewing every chunk. Tell me exactly which +Continuum inputs and controls to use, what must stay fixed between Queue runs, +and what output length I should expect after each action. Do not use older V3.7 UI. +``` + +AI products, plans, and repository-reading behavior change frequently. If a URL is not read correctly, paste or upload this README instead. Do not send private workflows, local paths, tokens, or credentials to a public AI service. The README and the UI remain the source of truth; an AI summary can still be wrong. + +## V3.8 supported surface + +The V3.8X2 baseline registers ten node IDs, including two deprecated loader compatibility IDs; V3.9 adds a separate Sampler and Reference Images helper. Current official workflows use Core Load Audio and Core Load Video plus the Video Adapter. + +- **H3 Continuum Sampler V3.8** — the Main sampler +- **H3 Continuum Finalize** — decoded Video/Audio assembly with optional seam handling +- **H3 Continuum Load Image** +- **H3 Continuum Load Audio** (legacy compatibility; new workflows use Core Load Audio) +- **H3 Continuum Load Video** (legacy compatibility; new workflows use Core Load Video + Video Adapter) +- **H3 Continuum Video Adapter** — Core VIDEO to H3 IMAGE/AUDIO, default 24fps, no Enable +- **H3 Continuum Second Pass** — the Advanced bridge for external latent processing or upscaling +- **H3 Continuum Reference Audios** — an ordered Reference Audio 1/2/3 bundle helper +- **H3 Continuum Reference Images** — optional additional Reference Images 4–9 +- **Decode Cache Helper** — the existing optional decoded-result cache + +The current frontend uses `Show Advanced Settings` / `Hide Advanced Settings` to change presentation without changing generation values. Saved drafts that still contain the former frontend-only `H3 Continuum View` property are migrated to the matching collapsed/expanded view and the legacy property is removed. If the frontend extension is unavailable, the complete Python-defined interface remains visible and executable. + +The official V3.8X2 workflow uses the Helper path: + +```text +H3 Continuum Sampler V3.8 + -> Decode Cache Helper + -> H3 Continuum Finalize + -> Create Video + -> Save Video +``` + +The Helper delegates cache MISS entries to native Core Video/Audio Decode. A manually wired direct Core Decode path remains supported, but it has no decoded-result cache. Save and upscaling remain external. Hi-Res Fix is not part of the V3.8X2 standard workflow; use Second Pass as the bridge when an external latent processor or upscaler is needed. External processors may change only Video LATENT spatial geometry while preserving physical groups, B/C/T, finite values, first-pass Audio, and the original Assembly Plan. Finalize accepts public `IMAGE`/`AUDIO` outputs and does not require a specific decoder class. SageAttention, Sol-Attn, and Spectrum remain external MODEL wrappers. See the [V3.8 Open Integration Contract](V38_OPEN_INTEGRATION_CONTRACT.md). + +## Complete V3.8 UI reference + +This section covers every user-facing control and socket on the seven public V3.8 nodes. Names such as `Power Lora Loader (rgthree)`, Spectrum, SageAttention, Core VAE Decode, and Core Save Video belong to ComfyUI or another extension. They may be used in a supplied workflow, but they are not Continuum controls. + +### H3 Continuum Sampler V3.8: required graph inputs + +| Input | What it does | Normal use | +|---|---|---| +| `model` | MiniMax H3 diffusion model, including any upstream MODEL wrappers or selected LoRA | Required | +| `clip` | MiniMax H3 text encoder | Required | +| `video_vae` | Encodes connected image conditioning; Continuum does not use it for final decoding | Required by the node; unused by pure T2VA | +| `sampler` | ComfyUI sampling algorithm | Required | +| `sigmas` | Noise schedule and effective step count | Required | +| `Sequence Prompt` | Complete text, list, timeline, or JSON sequence prompt | Required | + +The Sampler returns six outputs: a list of `video_latents`, a list of `audio_latents`, one `assembly_plan`, a text `status`, the selected `driving_audio` when used, and a `refine_context` for the Advanced Second Pass. Raw latent lists normally go through Core Video/Audio Decode before Finalize. + +### H3 Continuum Sampler V3.8: optional media inputs + +| Input | Role | Important distinction | +|---|---|---| +| `first_frame` | First Image for I2VA or FL2VA | Can also supply the output aspect ratio when `Size Source = First Image` | +| `last_frame` | Optional final-image constraint for FL2VA | Can trigger a Long Terminal Merge; the final pair may become one atomic review unit | +| `reference_image_1`–`reference_image_3` | Ordered appearance, identity, subject, or scene references | They never become the implicit size source | +| `Reference Images (Optional)` | Additional images 4–9 bundled by `H3 Continuum Reference Images` | Empty inputs are skipped; encoding remains in the Sampler | +| `Video Guide Frames` | A video loader's IMAGE frame batch, applied as a persistent guide | It does not carry the source video's audio; frames are interpreted at 24 fps | +| `Driving Audio` + `Driving Audio VAE` | Original audio timeline used as native guide conditioning | The selected source audio becomes final audio; generated audio and Audio Seam are bypassed | +| `Reference Audio (Optional)` + `Reference Audio VAE (Optional)` | Legacy single conditioning-only audio reference | Generated audio remains final audio | +| `Audio References (Optional)` | Ordered bundle from `H3 Continuum Reference Audios` | Do not connect this together with the legacy single Reference Audio path | +| `Still Image Guide (Optional)` | Compatibility socket inherited from the V3.7 guide contract | Experimental; not part of the V3.8X2 official workflow | + +#### First Image, Last Image, and Reference Images + +![First Image enabled and Last Image bypassed](images/v38-manual/first-last-image-bypass.png) + +- **First Image** establishes the opening image and is the normal visual starting point for I2VA/FL2VA. It is also the only media input that can drive `Size Source = First Image`. +- **Last Image** constrains the sequence ending. Leave it `OFF` for T2VA and ordinary I2VA. With a connected Last Image, extending or regenerating the sequence may rebuild the terminal pair. +- **Reference Images 1–9** guide identity or appearance throughout generation. Inputs 1–3 remain on the Sampler; optional inputs 4–9 are on `H3 Continuum Reference Images`, connected to the Sampler's `Reference Images (Optional)` socket. The helper only bundles images: resizing, hashing, and VAE encoding remain in the existing Sampler path. Empty slots are skipped; Picture numbers follow active slot order after any First/Last Image. Existing three-loader templates still work unchanged. + +Saved workflows with the earlier direct image 4/5 inputs migrate those links into a bundle after graph loading. Migration preserves the original source links and only removes a legacy input after its new link is verified. If a source is unavailable or its destination is occupied, the legacy link is retained with a warning. Old API prompts remain accepted; supplying both legacy and bundle images for the same slot is an explicit input conflict. + +![Three optional Reference Image loaders](images/v38-manual/reference-images-three-bypass.png) + +Every `H3 Continuum Load Image` has `Enable Image`. `ON` loads through ComfyUI Core. `OFF` uses native node bypass and makes that optional path behave as unconnected without deleting the node or cable. First, Last, and up to nine Reference Image inputs can therefore be connected without forcing every input to be active. An empty image bundle returns no reference and does not change the zero-to-three-image sampling/reuse contract. + +| Public loader | Controls | Outputs | +|---|---|---| +| `H3 Continuum Load Image` | `Enable Image`, `image` file | `IMAGE`, `MASK` | +| `H3 Continuum Load Audio` | `Enable Audio`, `audio` file | `AUDIO` | +| `H3 Continuum Load Video` | `Enable Video`, `Video` file, `Force Rate` | `images`, `audio` | + +#### Video, audio, and their bypass switches + +![Continuum Video and Audio loaders with Enable controls](images/v38-manual/input-video-audio-bypass.png) + +- `H3 Continuum Load Video`: `Enable Video`, `Video`, and `Force Rate`. `Force Rate = 0` uses source FPS; a positive value drops/duplicates frames to that rate while preserving nominal duration and audio. Connect `images` to `Video Guide Frames` when visual video guidance is wanted. Connect `audio` to `Driving Audio` only when the source audio should guide and become the final output. +- `H3 Continuum Load Audio`: `Enable Audio` and the audio file. Use it for Driving Audio or Reference Audio according to the socket you connect. +- The node titled `Audio Switch` in the example image is ComfyUI Core's `If/Else Switch` (`ComfySwitchNode`); `Fast Groups Bypasser (rgthree)` belongs to the rgthree extension. Neither is a Continuum node or a requirement of Continuum. The Continuum loaders already provide their own native ON/OFF controls. + +`Audio Continuity` is different from both audio inputs: it passes the previously generated audio context into the next generated chunk. It does not import a source track. Driving Audio replaces final generated audio; Reference Audio only conditions generation. + +### Main Sampler controls + +| Displayed control | Meaning | Recommended starting point | +|---|---|---| +| `Prompt Format` | `Auto`, `Fixed`, `List`, or `Timeline` interpretation of `Sequence Prompt` | `Auto` unless you need to force one parser | +| `Continuity` | Prior Video context retained at each chunk boundary | `Balanced — 22 frames` for Standard audiovisual continuation | +| `Base Seed` | Base for deterministic per-chunk seed derivation | Keep the value unchanged when continuing, comparing, or selecting Takes | +| `Control After Generate` | What ComfyUI does to `Base Seed` after a Queue | `fixed` for `Review Each Chunk`, resume, and controlled comparisons | +| `Audio Continuity` | Carries generated-audio context across boundaries | `true` for normal audiovisual generation | +| `Chunks` | Planned total chunk count, from 1 to 16 | Enter the final total, not the number to add next | +| `Seconds per Chunk` | Duration shared by every chunk | 5–15 seconds is the validated range; see duration notes below | +| `Total Length` | Read-only `Chunks × Seconds per Chunk` | Check only; it never edits either input | +| `Size Source` | `First Image` or `Manual` output geometry | First Image for I2VA/FL2VA; Manual for T2VA or exact dimensions | +| `Resolution` | First-Image sizing preset | `Draft — 0.30 MP` for tests, `Balanced — 0.60 MP` for more detail, or `Native 768` for the H3-native short edge | +| `Custom MP` | Custom First-Image pixel budget | Visible only when `Resolution = Custom` | +| `Width` / `Height` | Exact output canvas in 32-pixel steps | Visible/editable only in Manual mode; 32–16384 is accepted | +| `Run` | `Generate Full Video` or `Review Each Chunk` | Use Review when you want a human decision after each physical group | +| `Progress` | Saved raw chunks, safe resume, and Take history | Review turns it On automatically; enable it for resume/regeneration | +| `Ready to Queue` | Plain-language summary of the next Queue | Read it before pressing ComfyUI's top-right blue `Run` button | +| `Show Advanced Settings` / `Hide Advanced Settings` | Shows or hides technical controls | Visibility only; it does not change their values | + +`Continuity = Balanced — 22 frames` with `Continuation Backend = Standard` and `Audio Continuity = true` uses the current Masked AV route. `Fast — 5 frames`, `Strong — 39 frames (Experimental)`, and `Auto — conservative` safely resolve to the older Reference Context route while generated-audio continuity is on. With Audio Continuity off, Standard uses the Video-only masked route. + +#### Choose First Image sizing or Manual sizing + +![First Image sizing with Draft 0.30 MP](images/v38-manual/sampler-size-first-image-draft.png) + +`Size Source = First Image` preserves the connected First Image aspect ratio. `Resolution` then chooses its pixel area: + +| Preset | Target | +|---|---| +| `Draft — 0.30 MP` | Fastest and lowest-memory First Image starting point | +| `Balanced — 0.60 MP` | More detail with greater memory and processing cost | +| `Native 768` | 768 px short-edge target with a 1344 px long-edge cap | +| `Custom` | `Custom MP` selected by the user; separate from the Native 768 cap | + +![Size Source First Image tooltip](images/v38-manual/size-source-first-image-tooltip.png) + +`Size Source = Manual` ignores the preset and uses exact Width/Height values. Both must be multiples of 32. Manual is the normal choice for T2VA and also supports a 32×32 diagnostic canvas when only generated audio is being checked. + +![Manual Size Source with explicit Width and Height](images/v38-manual/size-source-manual-640.png) + +If First Image mode is selected but no usable First Image reaches the Sampler, V3.8 uses the stored Manual Width/Height as a safe fallback and reports the dimensions in `status`. Select `Size Source = Manual` to inspect or edit those values; Width/Height are hidden in the First Image facade. Reference Images and Video Guide never become implicit size sources. The historical `Auto / Landscape / Portrait / Square` values are migration/API compatibility values, not current Main choices. + +Size selection is built into the V3.8 Sampler. A separate megapixel, empty-latent, or image-size node is not required for the normal Continuum workflow. + +#### Choose a chunk duration + +- **5 seconds** is a valid H3 duration and the fastest practical choice for workflow checks, Review UI checks, and quick Takes. +- **8–10 seconds** usually reduces the number of handoffs while keeping each Queue easier to review and rerun. +- **15 seconds** is the [upstream MiniMax H3 native maximum per generation](https://github.com/MiniMax-AI/cli/blob/main/skill/h3-video/SKILL.md) and minimizes the number of boundaries, but it increases per-chunk time and memory. Longer is not automatically higher quality. +- Continuum accepts 4–30 seconds for compatibility and experimentation. Values above 15 seconds are not the same as an upstream native-duration recommendation and can be expensive at high resolution. + +For an audio-only listening diagnostic, use T2VA, `Size Source = Manual`, and a very small canvas such as 32×32, then play the saved video and ignore its picture. This is useful for hearing music or boundary continuity; it is not evidence of normal-resolution video quality. + +### Advanced Sampler controls + +| Control | When it appears | Meaning | +|---|---|---| +| `Continuation Backend` | Advanced open | `Standard` is the V3.8 path; `Compatibility` restores the older Reference Context route for comparison | +| `Regenerate From` | Advanced open and Progress On | `Auto` resumes the longest compatible prefix; `Chunk N` reuses earlier chunks and rebuilds N through the target | +| `Variation Nonce` | Explicit `Regenerate From = Chunk N` | `0`: automatic variation selection with interrupted-run resume; `1` or higher: a fixed variation value. Base Seed and the other generation settings stay unchanged | +| `Run Name (Optional Override)` | Advanced open and Progress On | Stable name for resume and Render History; blank uses the Sampler's automatic identity | +| `Reference Image Size` | Advanced open and a Reference Image connected | `Match Output` is practical; `Max Identity` preserves more reference detail and may use more memory | +| `Video Guide Size` | Advanced open and Video Guide connected | `Efficient — 0.4 MP`, `Balanced — 0.6 MP`, or `Match Output` | + +Frontend-managed IDs, selected Take IDs, one-shot review actions, diagnostics, preview, and legacy compatibility values are deliberately not editable as ordinary node widgets. The corresponding settings or action buttons below are the supported interface. + +### Review-state controls + +When `Run = Review Each Chunk`, `Progress` is On, and a review unit is ready, the settings view changes to this action view: + +| Button | What the next top-right ComfyUI `Run` does | +|---|---| +| `Use it and continue` | Accepts the current result, reuses it, and generates one next physical group | +| `Try this chunk again` | Keeps the earlier accepted prefix and creates another Take of the current review unit | +| `Start again from Chunk 1` | Selects a fresh branch from Chunk 1 with current settings and automatic variation; press Queue afterward. Saved Takes are kept. The selection resets to Auto after Queue | +| `Use it and finish the rest` | Accepts the current result and generates every remaining group without further review pauses | +| `Back to Settings` | Shows the normal settings; does not queue, reset, or discard the review | +| `Return to Review` | Returns from settings to the pending review; does not undo edited values | +| `Render History — N Takes` | Opens or closes stored Take browsing; N is Take count, not chunk count | +| `Previous Take` / `Next Take` | Changes the selected stored Take only | +| `Use This Take` | Makes the selected Take canonical on the next Queue; no new sampling is required | +| `Continue From Here` | Branches from the selected Take and generates the next review unit | + +The action button selects intent and gains a `✓`; it does not start generation. Press ComfyUI's top-right blue `Run` button afterward. Full step-by-step examples begin in [Review, continue, and revise a chunked video](#review-continue-and-revise-a-chunked-video). + +### H3 Continuum Reference Audios + +`H3 Continuum Reference Audios` bundles up to three standalone audio references behind one `Audio References (Optional)` Sampler socket. Connect the references without gaps and use `