demo.git - Vide-coded 64k demo system

Age	Commit message (Expand)	Author
15 hours	fix(cnn_v3/tools): zoom channel via WebGPU re-render + fill panel width	skal
15 hours	feat(cnn_v3/tools): click vignette to show enlarged channel below grid	skal
15 hours	fix(cnn_v3/tools): remove unused sampler binding from FULL_PACK_SHADER	skal
15 hours	feat(cnn_v3): add G-buffer visualizer + web sample loader (Phase 7)	skal
18 hours	fix(cnn_v3/tools): fix getElementById illegal invocation + auto-preload weights	skal
18 hours	fix(cnn_v3): fix texture format mismatches in cnn_v3_test sequence	skal
19 hours	feat(cnn_v3): wire trained weights into CNNv3Effect + add timeline test sequence	skal
19 hours	feat(cnn_v3): add weight assets to assets.txt, update HOW_TO_CNN export docs	skal
19 hours	docs(cnn_v3): add uv inline deps to export_cnn_v3_weights.py	skal
19 hours	fix size	skal
19 hours	feat(cnn_v3): patch alignment search, resume, Ctrl-C save	skal
19 hours	normalize sample dimension	skal
20 hours	docs(cnn_v3): add uv inline deps to train_cnn_v3.py + HOW_TO_CNN note	skal
20 hours	perf(cnn_v3): cache dataset images at init to avoid per-patch disk I/O	skal
21 hours	docs(cnn_v3): add Windows 10 + CUDA training section to HOW_TO_CNN §2	skal
21 hours	fix(cnn_v3): correct weight budget in docstring (3.9→5.4 KB f16)	skal
21 hours	fix(cnn_v3): resize target to albedo dims when sizes differ	skal
21 hours	docs(cnn_v3): add full Old House example to HOW_TO_CNN §1b	skal
21 hours	fix(cnn_v3): native OPEN_EXR_MULTILAYER + quiet render + flexible channel names	skal
22 hours	docs(cnn_v3): blender4 alias + Blender 4.5 LTS requirement for training data	skal
22 hours	fix(blender_export): version detection + Blender 5.x warning + cleanup	skal
22 hours	fix(cnn_v3): blender_export Blender 5 compositor activation + document Render...	skal
23 hours	feat(cnn_v3): blender_export print pack_blender_sample.py batch command after...	skal
23 hours	fix(cnn_v3): blender_export fallback socket name aliases for Shadow etc.	skal
23 hours	fix(cnn_v3): blender_export discard dir next to --output, not in /tmp	skal
23 hours	fix(cnn_v3): blender_export.py Blender 5 File Output node slots + file_name	skal
23 hours	docs(cnn_v3): clarify --output is a base dir, not a frame_### pattern	skal
23 hours	docs(cnn_v3): update HOW_TO_CNN for Blender 5.x compatibility	skal
23 hours	fix(cnn_v3): blender_export.py Blender 5.x API compatibility	skal
23 hours	fix(cnn_v3): blender_export --view-layer flag + fallback to layer[0]	skal
24 hours	feat(cnn_v3): gen_sample tool + 7 simple training samples	skal
24 hours	feat(cnn_v3): gen_sample tool + 7 simple training samples	skal
24 hours	feat(cnn_v3): gen_sample tool + 7 simple training samples	skal
24 hours	feat(cnn_v3): gen_sample tool + 7 simple training samples	skal
41 hours	refactor(cnn_v3): code review — comments, simplifications, test fix	skal
45 hours	feat(cnn_v3): HTML WebGPU tool (index.html + shaders.js + tester.js)	skal
45 hours	feat(cnn_v3): export script + HOW_TO_CNN.md playbook	skal
45 hours	feat(cnn_v3): Phase 6 — training script (train_cnn_v3.py + cnn_v3_utils.py)	skal
46 hours	docs(cnn_v3): update CNN_V3.md + HOWTO.md to reflect Phases 1-5 complete	skal
46 hours	feat(cnn_v3): Phase 5 complete — parity validation passing (36/36 tests)	skal
47 hours	feat(cnn_v3): Phase 4 complete — CNNv3Effect C++ + FiLM uniform upload	skal
47 hours	feat(cnn_v3): Phase 3 complete — WGSL U-Net inference shaders	skal
3 days	feat(cnn_v3): Phase 1 complete - GBufferEffect integrated + HOWTO playbook	skal
3 days	feat(cnn_v3): G-buffer phase 1 + training infrastructure	skal
3 days	docs(cnn_v3): full design doc — U-Net + FiLM architecture plan	skal
2026-03-05	add training photos	skal
2026-02-27	remove old files, add new training set	skal
2026-02-15	feat(cnn): add CNN v3 directory structure with training data	skal