Loading dataset... CachedTrajectoryDataset: 64 demos, 2048 samples Source: {'cache_root': '/data/libero/ood_objpos_task0_train', 'benchmark': 'libero_spatial', 'task_ids': [0], 'frame_stride': 3} Total: 2048 samples ✓ Train: 1946 samples ✓ Val: 102 samples Initializing model (type=para)... Loading DINOv2 model... ✓ DINOv2 backbone is trainable ✓ DINO embedding dim: 384 ✓ Learnable start keypoint embedding (dim=384) ✓ Feature convs: 3× Conv2d(3×3) at pred_size=64 ✓ Volume head → (B, 4, 32, 64, 64) ✓ Gripper head → (B, 4, 2, 64, 64) [1×1 conv, CE 2-class] ✓ Rotation head → (B, 4, 3, 32, 64, 64) [1×1 conv, CE] Trainable parameters: 32,880,520 / 32,880,520 (100.00%) Computing dataset stats from random subset: 500/2048 samples (seed=42) /data2/cameron/miniconda3/envs/uva/lib/python3.10/site-packages/wandb/sdk/data_types/image.py:324: DeprecationWarning: 'mode' parameter is deprecated and will be removed in Pillow 13 (2026-10-15) ✓ Saved stats cache: /data/cameron/para_normalized_losses/libero/checkpoints/para_ood_objpos2/dataset_stats.json ✓ Height range from dataset: [0.918972, 1.177174] m ✓ Gripper range from dataset: [-1.000000, 1.000000] ✓ Rotation range (delta rotvec): ['-0.019', '-0.108', '-0.003'] .. ['0.014', '0.081', '0.002'] ✓ Reference rotation: ['0.9994', '-0.0003', '-0.0339', '-0.0022'] ✓ Position range from dataset: ['-0.211', '-0.163', '0.919'] .. ['0.133', '0.055', '1.177'] Starting training for 9999 epochs... ✓ Rotation loss SKIPPED ✓ EMA loss weights initialized (vol=11.8, rot=3.5, grip=0.69) self._image = pil_image.fromarray( Epochs: 0%| | 0/9999 [00:00