/home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning) /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torch/functional.py:478: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:2894.) return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined] Image size: 480 loading dataset refcocog into memory... creating index... index created. DONE (t=6.45s) lavt_one Window size 12! Randomly initialize Multi-modal Swin Transformer weights. /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torchvision/transforms/functional.py:417: UserWarning: Argument 'interpolation' of type int is deprecated since 0.13 and will be removed in 0.15. Please use InterpolationMode enum. warnings.warn( /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torchvision/transforms/functional.py:417: UserWarning: Argument 'interpolation' of type int is deprecated since 0.13 and will be removed in 0.15. Please use InterpolationMode enum. warnings.warn( /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torchvision/transforms/functional.py:417: UserWarning: Argument 'interpolation' of type int is deprecated since 0.13 and will be removed in 0.15. Please use InterpolationMode enum. warnings.warn( /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torchvision/transforms/functional.py:417: UserWarning: Argument 'interpolation' of type int is deprecated since 0.13 and will be removed in 0.15. Please use InterpolationMode enum. warnings.warn( Test: [ 0/2573] eta: 2:27:43 time: 3.4448 data: 1.1081 max mem: 1021 Test: [ 100/2573] eta: 0:07:45 time: 0.1553 data: 0.0019 max mem: 1021 Test: [ 200/2573] eta: 0:06:55 time: 0.1594 data: 0.0017 max mem: 1021 Test: [ 300/2573] eta: 0:06:23 time: 0.1637 data: 0.0016 max mem: 1021 Test: [ 400/2573] eta: 0:06:01 time: 0.1643 data: 0.0017 max mem: 1021 Test: [ 500/2573] eta: 0:05:42 time: 0.1610 data: 0.0017 max mem: 1021 Test: [ 600/2573] eta: 0:05:25 time: 0.1616 data: 0.0017 max mem: 1021 Test: [ 700/2573] eta: 0:05:09 time: 0.1610 data: 0.0017 max mem: 1021 Test: [ 800/2573] eta: 0:04:52 time: 0.1702 data: 0.0017 max mem: 1021 Test: [ 900/2573] eta: 0:04:35 time: 0.1617 data: 0.0017 max mem: 1021 Test: [1000/2573] eta: 0:04:18 time: 0.1618 data: 0.0017 max mem: 1021 Test: [1100/2573] eta: 0:04:02 time: 0.1623 data: 0.0017 max mem: 1021 Test: [1200/2573] eta: 0:03:45 time: 0.1617 data: 0.0016 max mem: 1021 Test: [1300/2573] eta: 0:03:29 time: 0.1617 data: 0.0016 max mem: 1021 Test: [1400/2573] eta: 0:03:12 time: 0.1584 data: 0.0017 max mem: 1021 Test: [1500/2573] eta: 0:02:55 time: 0.1628 data: 0.0017 max mem: 1021 Test: [1600/2573] eta: 0:02:39 time: 0.1711 data: 0.0017 max mem: 1021 Test: [1700/2573] eta: 0:02:23 time: 0.1580 data: 0.0016 max mem: 1021 Test: [1800/2573] eta: 0:02:06 time: 0.1667 data: 0.0017 max mem: 1021 Test: [1900/2573] eta: 0:01:49 time: 0.1574 data: 0.0017 max mem: 1021 Test: [2000/2573] eta: 0:01:33 time: 0.1618 data: 0.0017 max mem: 1021 Test: [2100/2573] eta: 0:01:17 time: 0.1664 data: 0.0017 max mem: 1021 Test: [2200/2573] eta: 0:01:00 time: 0.1665 data: 0.0016 max mem: 1021 Test: [2300/2573] eta: 0:00:44 time: 0.1663 data: 0.0017 max mem: 1021 Test: [2400/2573] eta: 0:00:28 time: 0.1534 data: 0.0016 max mem: 1021 Test: [2500/2573] eta: 0:00:11 time: 0.1630 data: 0.0019 max mem: 1021 Test: Total time: 0:06:59 Final results: Mean IoU is 65.76 precision@0.5 = 74.16 precision@0.6 = 69.46 precision@0.7 = 63.17 precision@0.8 = 52.31 precision@0.9 = 26.49 overall IoU = 63.22 /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning) /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torch/functional.py:478: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:2894.) return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined] Image size: 480 Easy & Hard Example Experiments - dataset : refcocog, split : static loading dataset refcocog into memory... creating index... index created. DONE (t=6.49s) lavt_one Window size 12! Randomly initialize Multi-modal Swin Transformer weights. /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torchvision/transforms/functional.py:417: UserWarning: Argument 'interpolation' of type int is deprecated since 0.13 and will be removed in 0.15. Please use InterpolationMode enum. warnings.warn( /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torchvision/transforms/functional.py:417: UserWarning: Argument 'interpolation' of type int is deprecated since 0.13 and will be removed in 0.15. Please use InterpolationMode enum. warnings.warn( /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torchvision/transforms/functional.py:417: UserWarning: Argument 'interpolation' of type int is deprecated since 0.13 and will be removed in 0.15. Please use InterpolationMode enum. warnings.warn( /home/chaeyun/.conda/envs/cris/lib/python3.9/site-packages/torchvision/transforms/functional.py:417: UserWarning: Argument 'interpolation' of type int is deprecated since 0.13 and will be removed in 0.15. Please use InterpolationMode enum. warnings.warn( Test: [ 0/151] eta: 0:08:25 time: 3.3473 data: 1.0557 max mem: 1021 Test: [100/151] eta: 0:00:06 time: 0.0863 data: 0.0017 max mem: 1021 Test: Total time: 0:00:16 Final results: Mean IoU is 73.23 precision@0.5 = 83.44 precision@0.6 = 80.13 precision@0.7 = 76.82 precision@0.8 = 72.19 precision@0.9 = 39.07 overall IoU = 70.52