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>

src/training/train_gtauav.py

@@ -73,6 +73,7 @@ class TrainConfigGTAUAV:
     lrsclip_path: str = _LRSCLIP
     init_gate: float = 0.7
     baseline_mode: bool = False
+    shared_encoder: bool = True  # single DINOv3 WEB for both branches (saves ~4-5 GB)
 
     # Training.
     resume_from: str | None = None  # path to checkpoint for resuming
@@ -341,13 +342,15 @@ def train(cfg: TrainConfigGTAUAV) -> None:
         start_epoch = resume_ckpt.get("epoch", -1) + 1
     else:
         mode_str = "baseline (no text)" if cfg.baseline_mode else "with text (L1/L2/L3)"
-        LOGGER.info("Building model — %s", mode_str)
+        enc_str = "shared DINOv3 WEB" if cfg.shared_encoder else "asymmetric (WEB + SAT)"
+        LOGGER.info("Building model — %s, %s", mode_str, enc_str)
         model = AsymmetricEncoder(
             dino_web_path=cfg.dino_web_path,
             dino_sat_path=cfg.dino_sat_path,
             lrsclip_path=cfg.lrsclip_path,
             init_gate=cfg.init_gate,
             baseline_mode=cfg.baseline_mode,
+            shared_encoder=cfg.shared_encoder,
             device=cfg.device,
         ).to(cfg.device)