pikaliov da2d2ea90e Switch to shared DINOv3 WEB encoder (saves ~4-5 GB VRAM) ...

- Single DINOv3 WEB for both drone and satellite branches (shared_encoder=True default)
- One set of MONA adapters instead of two: 7M trainable vs 14M
- Total params: 438M (was 748M), trainable: 10.6M (was 17.6M)
- Asymmetric mode still available via shared_encoder=False
- Add gradient accumulation (grad_accum_steps, --grad-accum CLI flag)
- Update model summary in README

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-04-21 21:25:46 +03:00

..

__init__.py

Initial commit: caption quality test on UAV-VisLoc

2026-04-17 00:04:46 +03:00

grad_monitor.py

Add ML diagnostics tooling (W&B, TensorBoard, Grad-CAM, profiler) and gin configs

2026-04-21 20:30:50 +03:00

gradcam.py

Add ML diagnostics tooling (W&B, TensorBoard, Grad-CAM, profiler) and gin configs

2026-04-21 20:30:50 +03:00

plot_metrics.py

Add seaborn/matplotlib metric plots, auto-generated after each eval

2026-04-21 19:54:18 +03:00

profiling.py

Add ML diagnostics tooling (W&B, TensorBoard, Grad-CAM, profiler) and gin configs

2026-04-21 20:30:50 +03:00

trackers.py

Add ML diagnostics tooling (W&B, TensorBoard, Grad-CAM, profiler) and gin configs

2026-04-21 20:30:50 +03:00

train_gtauav.py

Switch to shared DINOv3 WEB encoder (saves ~4-5 GB VRAM)

2026-04-21 21:25:46 +03:00

train.py

Rewrite: GatedFusion architecture + UAV-GeoLoc dataset

2026-04-17 17:13:00 +03:00