demo.git - Vide-coded 64k demo system

Age	Commit message (Expand)	Author
8 hours	docs(cnn_v3): add uv inline deps to train_cnn_v3.py + HOW_TO_CNN note	skal
8 hours	perf(cnn_v3): cache dataset images at init to avoid per-patch disk I/O	skal
8 hours	fix(cnn_v3): correct weight budget in docstring (3.9→5.4 KB f16)	skal
8 hours	fix(cnn_v3): resize target to albedo dims when sizes differ	skal
8 hours	docs(cnn_v3): add full Old House example to HOW_TO_CNN §1b	skal
9 hours	fix(cnn_v3): native OPEN_EXR_MULTILAYER + quiet render + flexible channel names	skal
9 hours	fix(blender_export): version detection + Blender 5.x warning + cleanup	skal
10 hours	fix(cnn_v3): blender_export Blender 5 compositor activation + document Render...	skal
10 hours	feat(cnn_v3): blender_export print pack_blender_sample.py batch command after...	skal
10 hours	fix(cnn_v3): blender_export fallback socket name aliases for Shadow etc.	skal
10 hours	fix(cnn_v3): blender_export discard dir next to --output, not in /tmp	skal
10 hours	fix(cnn_v3): blender_export.py Blender 5 File Output node slots + file_name	skal
11 hours	docs(cnn_v3): clarify --output is a base dir, not a frame_### pattern	skal
11 hours	fix(cnn_v3): blender_export.py Blender 5.x API compatibility	skal
11 hours	fix(cnn_v3): blender_export --view-layer flag + fallback to layer[0]	skal
11 hours	feat(cnn_v3): gen_sample tool + 7 simple training samples	skal
11 hours	feat(cnn_v3): gen_sample tool + 7 simple training samples	skal
12 hours	feat(cnn_v3): gen_sample tool + 7 simple training samples	skal
29 hours	refactor(cnn_v3): code review — comments, simplifications, test fix	skal
33 hours	feat(cnn_v3): export script + HOW_TO_CNN.md playbook	skal
33 hours	feat(cnn_v3): Phase 6 — training script (train_cnn_v3.py + cnn_v3_utils.py)	skal
33 hours	feat(cnn_v3): Phase 5 complete — parity validation passing (36/36 tests)	skal
2 days	feat(cnn_v3): G-buffer phase 1 + training infrastructure	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