diff --git a/.gitattributes b/.gitattributes
index 61df17eb366cf1c2177abbb5e31194142fd945ee..c45ddb50a869f1574c57997eedf434fb68188c11 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -36,3 +36,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 MedNeXt_Base/fold_0/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_0/progress.png filter=lfs diff=lfs merge=lfs -text
 nnUNet_Large/fold_0/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_0/progress.png filter=lfs diff=lfs merge=lfs -text
 MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/progress.png filter=lfs diff=lfs merge=lfs -text
+nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/progress.png filter=lfs diff=lfs merge=lfs -text
diff --git a/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/dataset.json b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/dataset.json
new file mode 100644
index 0000000000000000000000000000000000000000..998fada443cfc594f132bb17b98c2cc25192d896
--- /dev/null
+++ b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/dataset.json
@@ -0,0 +1,46 @@
+{
+    "channel_names": {
+        "0": "CT"
+    },
+    "labels": {
+        "background": 0,
+        "BA": 1,
+        "R-P1P2": 2,
+        "L-P1P2": 3,
+        "R-ICA-C6-C7": 4,
+        "R-M1": 5,
+        "L-ICA-C6-C7": 6,
+        "L-M1": 7,
+        "R-Pcom": 8,
+        "L-Pcom": 9,
+        "Acom": 10,
+        "R-A1A2": 11,
+        "L-A1A2": 12,
+        "R-A3": 13,
+        "L-A3": 14,
+        "3rd-A2": 15,
+        "3rd-A3": 16,
+        "R-M2": 17,
+        "R-M3": 18,
+        "L-M2": 19,
+        "L-M3": 20,
+        "R-P3P4": 21,
+        "L-P3P4": 22,
+        "R-VA": 23,
+        "L-VA": 24,
+        "R-SCA": 25,
+        "L-SCA": 26,
+        "R-AICA": 27,
+        "L-AICA": 28,
+        "R-PICA": 29,
+        "L-PICA": 30,
+        "R-AChA": 31,
+        "L-AChA": 32,
+        "R-OA": 33,
+        "L-OA": 34,
+        "R-ICA-C1-C5": 35,
+        "L-ICA-C1-C5": 36
+    },
+    "numTraining": 25,
+    "file_ending": ".nii.gz"
+}
\ No newline at end of file
diff --git a/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/dataset_fingerprint.json b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/dataset_fingerprint.json
new file mode 100644
index 0000000000000000000000000000000000000000..e278577ae2fadbf012a75ccc57cd8c1b9c4ef448
--- /dev/null
+++ b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/dataset_fingerprint.json
@@ -0,0 +1,268 @@
+{
+    "foreground_intensity_properties_per_channel": {
+        "0": {
+            "max": 1813.9854736328125,
+            "mean": 211.27015686035156,
+            "median": 192.95703125,
+            "min": -91.01630401611328,
+            "percentile_00_5": 66.00397491455078,
+            "percentile_99_5": 618.0081787109375,
+            "std": 107.36483764648438
+        }
+    },
+    "median_relative_size_after_cropping": 1.0,
+    "shapes_after_crop": [
+        [
+            243,
+            327,
+            284
+        ],
+        [
+            203,
+            347,
+            289
+        ],
+        [
+            182,
+            313,
+            281
+        ],
+        [
+            224,
+            431,
+            351
+        ],
+        [
+            184,
+            417,
+            332
+        ],
+        [
+            216,
+            369,
+            286
+        ],
+        [
+            208,
+            321,
+            258
+        ],
+        [
+            204,
+            338,
+            270
+        ],
+        [
+            196,
+            386,
+            295
+        ],
+        [
+            295,
+            334,
+            269
+        ],
+        [
+            311,
+            371,
+            266
+        ],
+        [
+            168,
+            433,
+            326
+        ],
+        [
+            211,
+            328,
+            247
+        ],
+        [
+            203,
+            390,
+            316
+        ],
+        [
+            182,
+            365,
+            306
+        ],
+        [
+            527,
+            430,
+            319
+        ],
+        [
+            202,
+            379,
+            286
+        ],
+        [
+            200,
+            387,
+            290
+        ],
+        [
+            205,
+            345,
+            275
+        ],
+        [
+            224,
+            351,
+            287
+        ],
+        [
+            204,
+            396,
+            338
+        ],
+        [
+            223,
+            366,
+            300
+        ],
+        [
+            206,
+            340,
+            271
+        ],
+        [
+            214,
+            316,
+            274
+        ],
+        [
+            212,
+            360,
+            275
+        ]
+    ],
+    "spacings": [
+        [
+            0.625,
+            0.5078129768371582,
+            0.5078129768371582
+        ],
+        [
+            0.699951171875,
+            0.45703125,
+            0.45703125
+        ],
+        [
+            0.75,
+            0.50390625,
+            0.50390625
+        ],
+        [
+            0.70001220703125,
+            0.40625,
+            0.40625
+        ],
+        [
+            0.75,
+            0.3984375,
+            0.3984375
+        ],
+        [
+            0.75,
+            0.484375,
+            0.484375
+        ],
+        [
+            0.75,
+            0.544921875,
+            0.544921875
+        ],
+        [
+            0.75,
+            0.50390625,
+            0.50390625
+        ],
+        [
+            0.75,
+            0.431640625,
+            0.431640625
+        ],
+        [
+            0.5,
+            0.498046875,
+            0.498046875
+        ],
+        [
+            0.5,
+            0.498046875,
+            0.498046875
+        ],
+        [
+            0.7000007629394531,
+            0.40234375,
+            0.40234375
+        ],
+        [
+            0.75,
+            0.552734375,
+            0.552734375
+        ],
+        [
+            0.75,
+            0.431640625,
+            0.431640625
+        ],
+        [
+            0.75,
+            0.431640625,
+            0.431640625
+        ],
+        [
+            0.29998779296875,
+            0.4375,
+            0.4375
+        ],
+        [
+            0.75,
+            0.482421875,
+            0.482421875
+        ],
+        [
+            0.75,
+            0.4609375,
+            0.4609375
+        ],
+        [
+            0.70001220703125,
+            0.4921875,
+            0.4921875
+        ],
+        [
+            0.70001220703125,
+            0.486328125,
+            0.486328125
+        ],
+        [
+            0.70001220703125,
+            0.419921875,
+            0.419921875
+        ],
+        [
+            0.699951171875,
+            0.451171875,
+            0.451171875
+        ],
+        [
+            0.70001220703125,
+            0.48828125,
+            0.48828125
+        ],
+        [
+            0.70001220703125,
+            0.482421875,
+            0.482421875
+        ],
+        [
+            0.70001220703125,
+            0.486328125,
+            0.486328125
+        ]
+    ]
+}
\ No newline at end of file
diff --git a/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/checkpoint_best.pth b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/checkpoint_best.pth
new file mode 100644
index 0000000000000000000000000000000000000000..abef8b1e738a1864216b90755293c0ebdcd42c6f
--- /dev/null
+++ b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/checkpoint_best.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa28d1eebcfcc63d5b7d6c052e8eabfbd86410633facef41549795bce4cee8f9
+size 819660547
diff --git a/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/checkpoint_latest.pth b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/checkpoint_latest.pth
new file mode 100644
index 0000000000000000000000000000000000000000..513bfd3d5dbf610464aedeb744535e865940bae7
--- /dev/null
+++ b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/checkpoint_latest.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4ef9e23aded21021a4137f60be272fcdca92a1d8615296394cbfb7446aa2219
+size 819680811
diff --git a/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/debug.json b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/debug.json
new file mode 100644
index 0000000000000000000000000000000000000000..b4caf9610315323f9a8c29370177b67366b099de
--- /dev/null
+++ b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/debug.json
@@ -0,0 +1,55 @@
+{
+    "_best_ema": "None",
+    "batch_size": "2",
+    "configuration_manager": "{'data_identifier': 'nnUNetPlans_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [128, 256, 192], 'median_image_size_in_voxels': [214.0, 352.0, 281.0], 'spacing': [0.70001220703125, 0.482421875, 0.482421875], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': False}",
+    "configuration_name": "3d_fullres",
+    "cudnn_version": 91900,
+    "current_epoch": "0",
+    "dataloader_train": "<batchgenerators.dataloading.nondet_multi_threaded_augmenter.NonDetMultiThreadedAugmenter object at 0x7f07cd8712b0>",
+    "dataloader_train.generator": "<nnunetv2.training.dataloading.data_loader.nnUNetDataLoader object at 0x7f07cd871160>",
+    "dataloader_train.generator.transforms": "ComposeTransforms( transforms = [SpatialTransform( patch_size = [128, 256, 192], patch_center_dist_from_border = [0, 0, 0], random_crop = False, p_elastic_deform = 0, elastic_deform_scale = (0, 0.2), elastic_deform_magnitude = (0, 0.2), p_rotation = 0.2, rotation = (-0.5235987755982988, 0.5235987755982988), p_rot_per_axis = 1, p_scaling = 0.2, scaling = (0.7, 1.4), p_synchronize_scaling_across_axes = 1, p_synchronize_def_scale_across_axes = 0, bg_style_seg_sampling = False, seg_tiebreak = 'nearest', mode_seg = 'bilinear', border_mode_seg = 'constant', center_deformation = True, mode_image = 'bilinear', padding_mode_image = 'zeros', padding_value_seg = -1, padding_value_image = 0, align_corners = False, _grid_cache = {} ), RandomTransform(p=0.1, transform=GaussianNoiseTransform( noise_variance = (0, 0.1), p_per_channel = 1, synchronize_channels = True )), RandomTransform(p=0.2, transform=GaussianBlurTransform( blur_sigma = (0.5, 1.0), benchmark = True, synchronize_channels = False, synchronize_axes = False, p_per_channel = 0.5, benchmark_use_fft = {}, benchmark_num_runs = 9 )), RandomTransform(p=0.15, transform=MultiplicativeBrightnessTransform( multiplier_range = BGContrast(contrast_range=(0.75, 1.25)), synchronize_channels = False, p_per_channel = 1 )), RandomTransform(p=0.15, transform=ContrastTransform( contrast_range = BGContrast(contrast_range=(0.75, 1.25)), preserve_range = True, synchronize_channels = False, p_per_channel = 1.0 )), RandomTransform(p=0.25, transform=SimulateLowResolutionTransform( scale = (0.5, 1), synchronize_channels = False, synchronize_axes = True, ignore_axes = None, allowed_channels = None, p_per_channel = 0.5, upmodes = {1: 'linear', 2: 'bilinear', 3: 'trilinear'} )), RandomTransform(p=0.1, transform=GammaTransform( gamma = BGContrast(contrast_range=(0.7, 1.5)), p_invert_image = 1.0, synchronize_channels = False, p_per_channel = 1.0, p_retain_stats = 1.0 )), RandomTransform(p=0.3, transform=GammaTransform( gamma = BGContrast(contrast_range=(0.7, 1.5)), p_invert_image = 0.0, synchronize_channels = False, p_per_channel = 1.0, p_retain_stats = 1.0 )), MirrorTransform( allowed_axes = (0, 1, 2) ), RemoveLabelTransform( segmentation_channels = None, label_value = -1, set_to = 0 ), DownsampleSegForDSTransform( ds_scales = [[np.float64(1.0), np.float64(1.0), np.float64(1.0)], [np.float64(0.5), np.float64(0.5), np.float64(0.5)], [np.float64(0.25), np.float64(0.25), np.float64(0.25)], [np.float64(0.125), np.float64(0.125), np.float64(0.125)], [np.float64(0.0625), np.float64(0.0625), np.float64(0.0625)]] )] )",
+    "dataloader_train.num_processes": "8",
+    "dataloader_train.transform": "None",
+    "dataloader_val": "<batchgenerators.dataloading.nondet_multi_threaded_augmenter.NonDetMultiThreadedAugmenter object at 0x7f07cd84a710>",
+    "dataloader_val.generator": "<nnunetv2.training.dataloading.data_loader.nnUNetDataLoader object at 0x7f07cd84a5d0>",
+    "dataloader_val.generator.transforms": "ComposeTransforms( transforms = [RemoveLabelTransform( segmentation_channels = None, label_value = -1, set_to = 0 ), DownsampleSegForDSTransform( ds_scales = [[np.float64(1.0), np.float64(1.0), np.float64(1.0)], [np.float64(0.5), np.float64(0.5), np.float64(0.5)], [np.float64(0.25), np.float64(0.25), np.float64(0.25)], [np.float64(0.125), np.float64(0.125), np.float64(0.125)], [np.float64(0.0625), np.float64(0.0625), np.float64(0.0625)]] )] )",
+    "dataloader_val.num_processes": "4",
+    "dataloader_val.transform": "None",
+    "dataset_json": "{'channel_names': {'0': 'CT'}, 'labels': {'background': 0, 'BA': 1, 'R-P1P2': 2, 'L-P1P2': 3, 'R-ICA-C6-C7': 4, 'R-M1': 5, 'L-ICA-C6-C7': 6, 'L-M1': 7, 'R-Pcom': 8, 'L-Pcom': 9, 'Acom': 10, 'R-A1A2': 11, 'L-A1A2': 12, 'R-A3': 13, 'L-A3': 14, '3rd-A2': 15, '3rd-A3': 16, 'R-M2': 17, 'R-M3': 18, 'L-M2': 19, 'L-M3': 20, 'R-P3P4': 21, 'L-P3P4': 22, 'R-VA': 23, 'L-VA': 24, 'R-SCA': 25, 'L-SCA': 26, 'R-AICA': 27, 'L-AICA': 28, 'R-PICA': 29, 'L-PICA': 30, 'R-AChA': 31, 'L-AChA': 32, 'R-OA': 33, 'L-OA': 34, 'R-ICA-C1-C5': 35, 'L-ICA-C1-C5': 36}, 'numTraining': 25, 'file_ending': '.nii.gz'}",
+    "device": "cuda:0",
+    "disable_checkpointing": "False",
+    "enable_deep_supervision": "True",
+    "fold": "1",
+    "folder_with_segs_from_previous_stage": "None",
+    "gpu_name": "NVIDIA RTX PRO 6000 Blackwell Server Edition",
+    "grad_scaler": "<torch.amp.grad_scaler.GradScaler object at 0x7f09d49b5d30>",
+    "hostname": "9caded48-fa6b-4114-a823-0c6d721b2ea8-frcrt",
+    "inference_allowed_mirroring_axes": "(0, 1, 2)",
+    "initial_lr": "0.01",
+    "is_cascaded": "False",
+    "is_ddp": "False",
+    "label_manager": "<nnunetv2.utilities.label_handling.label_handling.LabelManager object at 0x7f09d49b5be0>",
+    "local_rank": "0",
+    "log_file": "/marimo/TopBrain2026/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/training_log_2026_10_16_18_49_38.txt",
+    "logger": "<nnunetv2.training.logging.nnunet_logger.MetaLogger object at 0x7f09d49b5e80>",
+    "loss": "DeepSupervisionWrapper(\n  (loss): DC_and_CE_loss(\n    (ce): RobustCrossEntropyLoss()\n    (dc): MemoryEfficientSoftDiceLoss()\n  )\n)",
+    "lr_scheduler": "<nnunetv2.training.lr_scheduler.polylr.PolyLRScheduler object at 0x7f09d49b78c0>",
+    "my_init_kwargs": "{'plans': {'dataset_name': 'Dataset001_CT', 'plans_name': 'nnUNetResEncUNetLPlans', 'original_median_spacing_after_transp': [0.70001220703125, 0.482421875, 0.482421875], 'original_median_shape_after_transp': [206, 365, 286], 'image_reader_writer': 'SimpleITKIO', 'transpose_forward': [0, 1, 2], 'transpose_backward': [0, 1, 2], 'configurations': {'2d': {'data_identifier': 'nnUNetPlans_2d', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 70, 'patch_size': [384, 320], 'median_image_size_in_voxels': [352.0, 281.0], 'spacing': [0.482421875, 0.482421875], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 7, 'features_per_stage': [32, 64, 128, 256, 512, 512, 512], 'conv_op': 'torch.nn.modules.conv.Conv2d', 'kernel_sizes': [[3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3]], 'strides': [[1, 1], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm2d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': True}, '3d_fullres': {'data_identifier': 'nnUNetPlans_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [128, 256, 192], 'median_image_size_in_voxels': [214.0, 352.0, 281.0], 'spacing': [0.70001220703125, 0.482421875, 0.482421875], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': False}}, 'experiment_planner_used': 'nnUNetPlannerResEncL', 'label_manager': 'LabelManager', 'foreground_intensity_properties_per_channel': {'0': {'max': 1813.9854736328125, 'mean': 211.27015686035156, 'median': 192.95703125, 'min': -91.01630401611328, 'percentile_00_5': 66.00397491455078, 'percentile_99_5': 618.0081787109375, 'std': 107.36483764648438}}}, 'configuration': '3d_fullres', 'fold': 1, 'dataset_json': {'channel_names': {'0': 'CT'}, 'labels': {'background': 0, 'BA': 1, 'R-P1P2': 2, 'L-P1P2': 3, 'R-ICA-C6-C7': 4, 'R-M1': 5, 'L-ICA-C6-C7': 6, 'L-M1': 7, 'R-Pcom': 8, 'L-Pcom': 9, 'Acom': 10, 'R-A1A2': 11, 'L-A1A2': 12, 'R-A3': 13, 'L-A3': 14, '3rd-A2': 15, '3rd-A3': 16, 'R-M2': 17, 'R-M3': 18, 'L-M2': 19, 'L-M3': 20, 'R-P3P4': 21, 'L-P3P4': 22, 'R-VA': 23, 'L-VA': 24, 'R-SCA': 25, 'L-SCA': 26, 'R-AICA': 27, 'L-AICA': 28, 'R-PICA': 29, 'L-PICA': 30, 'R-AChA': 31, 'L-AChA': 32, 'R-OA': 33, 'L-OA': 34, 'R-ICA-C1-C5': 35, 'L-ICA-C1-C5': 36}, 'numTraining': 25, 'file_ending': '.nii.gz'}, 'device': device(type='cuda')}",
+    "network": "ResidualEncoderUNet",
+    "num_epochs": "1000",
+    "num_input_channels": "1",
+    "num_iterations_per_epoch": "250",
+    "num_val_iterations_per_epoch": "50",
+    "optimizer": "SGD (\nParameter Group 0\n    dampening: 0\n    differentiable: False\n    foreach: None\n    fused: None\n    initial_lr: 0.01\n    lr: 0.01\n    maximize: False\n    momentum: 0.99\n    nesterov: True\n    weight_decay: 3e-05\n)",
+    "output_folder": "/marimo/TopBrain2026/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1",
+    "output_folder_base": "/marimo/TopBrain2026/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres",
+    "oversample_foreground_percent": "0.33",
+    "plans_manager": "{'dataset_name': 'Dataset001_CT', 'plans_name': 'nnUNetResEncUNetLPlans', 'original_median_spacing_after_transp': [0.70001220703125, 0.482421875, 0.482421875], 'original_median_shape_after_transp': [206, 365, 286], 'image_reader_writer': 'SimpleITKIO', 'transpose_forward': [0, 1, 2], 'transpose_backward': [0, 1, 2], 'configurations': {'2d': {'data_identifier': 'nnUNetPlans_2d', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 70, 'patch_size': [384, 320], 'median_image_size_in_voxels': [352.0, 281.0], 'spacing': [0.482421875, 0.482421875], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 7, 'features_per_stage': [32, 64, 128, 256, 512, 512, 512], 'conv_op': 'torch.nn.modules.conv.Conv2d', 'kernel_sizes': [[3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3]], 'strides': [[1, 1], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm2d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': True}, '3d_fullres': {'data_identifier': 'nnUNetPlans_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [128, 256, 192], 'median_image_size_in_voxels': [214.0, 352.0, 281.0], 'spacing': [0.70001220703125, 0.482421875, 0.482421875], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': False}}, 'experiment_planner_used': 'nnUNetPlannerResEncL', 'label_manager': 'LabelManager', 'foreground_intensity_properties_per_channel': {'0': {'max': 1813.9854736328125, 'mean': 211.27015686035156, 'median': 192.95703125, 'min': -91.01630401611328, 'percentile_00_5': 66.00397491455078, 'percentile_99_5': 618.0081787109375, 'std': 107.36483764648438}}}",
+    "preprocessed_dataset_folder": "/marimo/TopBrain2026/nnUNet_preprocessed/Dataset001_CT/nnUNetPlans_3d_fullres",
+    "preprocessed_dataset_folder_base": "/marimo/TopBrain2026/nnUNet_preprocessed/Dataset001_CT",
+    "probabilistic_oversampling": "False",
+    "save_every": "50",
+    "torch_version": "2.11.0+cu130",
+    "was_initialized": "True",
+    "weight_decay": "3e-05"
+}
\ No newline at end of file
diff --git a/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/progress.png b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/progress.png
new file mode 100644
index 0000000000000000000000000000000000000000..51a2e58ab746c10b2fc28b00b88b4be54dc7fc73
--- /dev/null
+++ b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/progress.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b9ec21c5df5eec6258d71d84404092d5ba2ded6693cb649e689c8fced3f2eac
+size 965684
diff --git a/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/training_log_2026_10_16_18_49_38.txt b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/training_log_2026_10_16_18_49_38.txt
new file mode 100644
index 0000000000000000000000000000000000000000..c94d4c3bb32a030b59e6e4349f95515ee921ed37
--- /dev/null
+++ b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_1/training_log_2026_10_16_18_49_38.txt
@@ -0,0 +1,1246 @@
+
+#######################################################################
+Please cite the following paper when using nnU-Net:
+Isensee, F., Jaeger, P. F., Kohl, S. A., Petersen, J., & Maier-Hein, K. H. (2021). nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nature methods, 18(2), 203-211.
+#######################################################################
+ 
+2026-10-16 18:49:39.986004: do_dummy_2d_data_aug: False 
+2026-10-16 18:49:39.987213: Creating new 5-fold cross-validation split... 
+2026-10-16 18:49:40.027806: Desired fold for training: 1 
+2026-10-16 18:49:40.027980: This split has 20 training and 5 validation cases. 
+
+This is the configuration used by this training:
+Configuration name: 3d_fullres
+ {'data_identifier': 'nnUNetPlans_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [128, 256, 192], 'median_image_size_in_voxels': [214.0, 352.0, 281.0], 'spacing': [0.70001220703125, 0.482421875, 0.482421875], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': False} 
+ 
+These are the global plan.json settings:
+ {'dataset_name': 'Dataset001_CT', 'plans_name': 'nnUNetResEncUNetLPlans', 'original_median_spacing_after_transp': [0.70001220703125, 0.482421875, 0.482421875], 'original_median_shape_after_transp': [206, 365, 286], 'image_reader_writer': 'SimpleITKIO', 'transpose_forward': [0, 1, 2], 'transpose_backward': [0, 1, 2], 'experiment_planner_used': 'nnUNetPlannerResEncL', 'label_manager': 'LabelManager', 'foreground_intensity_properties_per_channel': {'0': {'max': 1813.9854736328125, 'mean': 211.27015686035156, 'median': 192.95703125, 'min': -91.01630401611328, 'percentile_00_5': 66.00397491455078, 'percentile_99_5': 618.0081787109375, 'std': 107.36483764648438}}} 
+ 
+2026-10-16 18:49:45.600356: Unable to plot network architecture: 
+2026-10-16 18:49:45.600570: No module named 'hiddenlayer' 
+2026-10-16 18:49:45.640082:  
+2026-10-16 18:49:45.640424: Epoch 0 
+2026-10-16 18:49:45.640959: Current learning rate: 0.01 
+2026-10-16 18:55:23.514946: train_loss 0.2591 
+2026-10-16 18:55:23.515968: val_loss 0.0296 
+2026-10-16 18:55:23.516478: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0)] 
+2026-10-16 18:55:23.516773: Epoch time: 337.88 s 
+2026-10-16 18:55:23.517028: Yayy! New best EMA pseudo Dice: 0.0 
+2026-10-16 18:55:24.949099:  
+2026-10-16 18:55:24.949759: Epoch 1 
+2026-10-16 18:55:24.950164: Current learning rate: 0.00999 
+2026-10-16 18:59:33.879441: train_loss 0.0187 
+2026-10-16 18:59:33.880078: val_loss 0.0134 
+2026-10-16 18:59:33.880706: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0)] 
+2026-10-16 18:59:33.881106: Epoch time: 248.93 s 
+2026-10-16 18:59:34.691124:  
+2026-10-16 18:59:34.693373: Epoch 2 
+2026-10-16 18:59:34.694135: Current learning rate: 0.00998 
+2026-10-16 19:03:51.919702: train_loss 0.0127 
+2026-10-16 19:03:51.920636: val_loss 0.01 
+2026-10-16 19:03:51.921087: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0)] 
+2026-10-16 19:03:51.921407: Epoch time: 257.23 s 
+2026-10-16 19:03:52.815713:  
+2026-10-16 19:03:52.816322: Epoch 3 
+2026-10-16 19:03:52.816696: Current learning rate: 0.00997 
+2026-10-16 19:08:00.931754: train_loss 0.0061 
+2026-10-16 19:08:00.932928: val_loss 0.0103 
+2026-10-16 19:08:00.933593: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0)] 
+2026-10-16 19:08:00.933933: Epoch time: 248.12 s 
+2026-10-16 19:08:01.733389:  
+2026-10-16 19:08:01.734126: Epoch 4 
+2026-10-16 19:08:01.734473: Current learning rate: 0.00996 
+2026-10-16 19:12:05.954229: train_loss 0.0034 
+2026-10-16 19:12:05.955235: val_loss 0.0015 
+2026-10-16 19:12:05.955903: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0066), np.float32(0.0)] 
+2026-10-16 19:12:05.956228: Epoch time: 244.22 s 
+2026-10-16 19:12:05.956517: Yayy! New best EMA pseudo Dice: 0.0 
+2026-10-16 19:12:09.369729:  
+2026-10-16 19:12:09.370600: Epoch 5 
+2026-10-16 19:12:09.370933: Current learning rate: 0.00995 
+2026-10-16 19:16:10.591295: train_loss -0.0003 
+2026-10-16 19:16:10.591928: val_loss -0.0037 
+2026-10-16 19:16:10.592645: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0734), np.float32(0.0)] 
+2026-10-16 19:16:10.593563: Epoch time: 241.22 s 
+2026-10-16 19:16:10.593848: Yayy! New best EMA pseudo Dice: 0.00019999999494757503 
+2026-10-16 19:16:11.888781:  
+2026-10-16 19:16:11.889869: Epoch 6 
+2026-10-16 19:16:11.890163: Current learning rate: 0.00995 
+2026-10-16 19:20:36.513997: train_loss -0.0061 
+2026-10-16 19:20:36.514472: val_loss -0.01 
+2026-10-16 19:20:36.514929: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1014), np.float32(0.0)] 
+2026-10-16 19:20:36.515182: Epoch time: 264.63 s 
+2026-10-16 19:20:36.515428: Yayy! New best EMA pseudo Dice: 0.0005000000237487257 
+2026-10-16 19:20:37.837101:  
+2026-10-16 19:20:37.837888: Epoch 7 
+2026-10-16 19:20:37.838336: Current learning rate: 0.00994 
+2026-10-16 19:24:37.357036: train_loss -0.0067 
+2026-10-16 19:24:37.357776: val_loss -0.0107 
+2026-10-16 19:24:37.358352: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1043), np.float32(0.0)] 
+2026-10-16 19:24:37.359164: Epoch time: 239.52 s 
+2026-10-16 19:24:37.359544: Yayy! New best EMA pseudo Dice: 0.000699999975040555 
+2026-10-16 19:24:38.664755:  
+2026-10-16 19:24:38.665408: Epoch 8 
+2026-10-16 19:24:38.665854: Current learning rate: 0.00993 
+2026-10-16 19:28:38.998289: train_loss -0.0089 
+2026-10-16 19:28:38.998851: val_loss -0.011 
+2026-10-16 19:28:38.999250: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1292), np.float32(0.0)] 
+2026-10-16 19:28:38.999918: Epoch time: 240.34 s 
+2026-10-16 19:28:39.000284: Yayy! New best EMA pseudo Dice: 0.0010000000474974513 
+2026-10-16 19:28:42.166009:  
+2026-10-16 19:28:42.166766: Epoch 9 
+2026-10-16 19:28:42.167146: Current learning rate: 0.00992 
+2026-10-16 19:33:01.868135: train_loss -0.0137 
+2026-10-16 19:33:01.868806: val_loss -0.016 
+2026-10-16 19:33:01.869474: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1667), np.float32(0.0103)] 
+2026-10-16 19:33:01.870169: Epoch time: 259.7 s 
+2026-10-16 19:33:01.870540: Yayy! New best EMA pseudo Dice: 0.00139999995008111 
+2026-10-16 19:33:03.159276:  
+2026-10-16 19:33:03.160081: Epoch 10 
+2026-10-16 19:33:03.160432: Current learning rate: 0.00991 
+2026-10-16 19:37:06.023160: train_loss -0.0174 
+2026-10-16 19:37:06.024091: val_loss -0.0206 
+2026-10-16 19:37:06.024452: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0939), np.float32(0.3073)] 
+2026-10-16 19:37:06.024988: Epoch time: 242.87 s 
+2026-10-16 19:37:06.025386: Yayy! New best EMA pseudo Dice: 0.002400000113993883 
+2026-10-16 19:37:07.358631:  
+2026-10-16 19:37:07.359434: Epoch 11 
+2026-10-16 19:37:07.359788: Current learning rate: 0.0099 
+2026-10-16 19:41:19.481375: train_loss -0.0235 
+2026-10-16 19:41:19.482236: val_loss -0.0237 
+2026-10-16 19:41:19.482739: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3618), np.float32(0.0)] 
+2026-10-16 19:41:19.483237: Epoch time: 252.12 s 
+2026-10-16 19:41:19.483728: Yayy! New best EMA pseudo Dice: 0.003100000089034438 
+2026-10-16 19:41:20.922831:  
+2026-10-16 19:41:20.923765: Epoch 12 
+2026-10-16 19:41:20.924215: Current learning rate: 0.00989 
+2026-10-16 19:45:33.092563: train_loss -0.0287 
+2026-10-16 19:45:33.093460: val_loss -0.0295 
+2026-10-16 19:45:33.094168: Pseudo dice [np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4044), np.float32(0.046)] 
+2026-10-16 19:45:33.094597: Epoch time: 252.17 s 
+2026-10-16 19:45:33.095045: Yayy! New best EMA pseudo Dice: 0.004100000020116568 
+2026-10-16 19:45:35.977889:  
+2026-10-16 19:45:35.978530: Epoch 13 
+2026-10-16 19:45:35.978909: Current learning rate: 0.00988 
+2026-10-16 19:49:27.356544: train_loss -0.0324 
+2026-10-16 19:49:27.357554: val_loss -0.0319 
+2026-10-16 19:49:27.358366: Pseudo dice [np.float32(0.1447), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1939), np.float32(0.4295)] 
+2026-10-16 19:49:27.358824: Epoch time: 231.38 s 
+2026-10-16 19:49:27.359647: Yayy! New best EMA pseudo Dice: 0.005799999926239252 
+2026-10-16 19:49:30.725897:  
+2026-10-16 19:49:30.726657: Epoch 14 
+2026-10-16 19:49:30.727059: Current learning rate: 0.00987 
+2026-10-16 19:53:40.225482: train_loss -0.0384 
+2026-10-16 19:53:40.225934: val_loss -0.0399 
+2026-10-16 19:53:40.226575: Pseudo dice [np.float32(0.2759), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0004), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4728), np.float32(1e-04)] 
+2026-10-16 19:53:40.227044: Epoch time: 249.5 s 
+2026-10-16 19:53:40.227525: Yayy! New best EMA pseudo Dice: 0.007300000172108412 
+2026-10-16 19:53:41.563872:  
+2026-10-16 19:53:41.564726: Epoch 15 
+2026-10-16 19:53:41.565294: Current learning rate: 0.00986 
+2026-10-16 19:57:38.538229: train_loss -0.0407 
+2026-10-16 19:57:38.538830: val_loss -0.0401 
+2026-10-16 19:57:38.539501: Pseudo dice [np.float32(0.245), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0004), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.522), np.float32(0.0012)] 
+2026-10-16 19:57:38.539870: Epoch time: 236.98 s 
+2026-10-16 19:57:38.540073: Yayy! New best EMA pseudo Dice: 0.008700000122189522 
+2026-10-16 19:57:41.839335:  
+2026-10-16 19:57:41.840040: Epoch 16 
+2026-10-16 19:57:41.840720: Current learning rate: 0.00986 
+2026-10-16 20:01:28.424091: train_loss -0.0473 
+2026-10-16 20:01:28.425125: val_loss -0.0441 
+2026-10-16 20:01:28.425718: Pseudo dice [np.float32(0.2831), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0025), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5146), np.float32(0.0002)] 
+2026-10-16 20:01:28.425976: Epoch time: 226.59 s 
+2026-10-16 20:01:28.426131: Yayy! New best EMA pseudo Dice: 0.010099999606609344 
+2026-10-16 20:01:29.858847:  
+2026-10-16 20:01:29.859586: Epoch 17 
+2026-10-16 20:01:29.859832: Current learning rate: 0.00985 
+2026-10-16 20:05:20.565019: train_loss -0.0518 
+2026-10-16 20:05:20.565799: val_loss -0.0534 
+2026-10-16 20:05:20.566486: Pseudo dice [np.float32(0.3424), np.float32(0.0), np.float32(0.0), np.float32(0.0284), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0101), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5021), np.float32(0.0062)] 
+2026-10-16 20:05:20.566889: Epoch time: 230.71 s 
+2026-10-16 20:05:20.567211: Yayy! New best EMA pseudo Dice: 0.011500000022351742 
+2026-10-16 20:05:22.342663:  
+2026-10-16 20:05:22.343364: Epoch 18 
+2026-10-16 20:05:22.343758: Current learning rate: 0.00984 
+2026-10-16 20:09:16.430993: train_loss -0.0562 
+2026-10-16 20:09:16.431998: val_loss -0.0583 
+2026-10-16 20:09:16.432643: Pseudo dice [np.float32(0.3729), np.float32(0.0), np.float32(0.0), np.float32(0.1662), np.float32(0.0), np.float32(0.0094), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(1e-04), np.float32(0.0), np.float32(0.0141), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.001), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5234), np.float32(0.143)] 
+2026-10-16 20:09:16.432880: Epoch time: 234.09 s 
+2026-10-16 20:09:16.433283: Yayy! New best EMA pseudo Dice: 0.013799999840557575 
+2026-10-16 20:09:20.117588:  
+2026-10-16 20:09:20.118398: Epoch 19 
+2026-10-16 20:09:20.118861: Current learning rate: 0.00983 
+2026-10-16 20:13:06.814720: train_loss -0.0632 
+2026-10-16 20:13:06.815690: val_loss -0.0606 
+2026-10-16 20:13:06.816252: Pseudo dice [np.float32(0.4664), np.float32(0.0), np.float32(0.0), np.float32(0.3811), np.float32(0.0), np.float32(0.0209), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0014), np.float32(0.0), np.float32(0.0762), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0181), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5023), np.float32(0.0002)] 
+2026-10-16 20:13:06.816834: Epoch time: 226.7 s 
+2026-10-16 20:13:06.817272: Yayy! New best EMA pseudo Dice: 0.016499999910593033 
+2026-10-16 20:13:08.411136:  
+2026-10-16 20:13:08.415418: Epoch 20 
+2026-10-16 20:13:08.415794: Current learning rate: 0.00982 
+2026-10-16 20:16:37.667553: train_loss -0.074 
+2026-10-16 20:16:37.668500: val_loss -0.0748 
+2026-10-16 20:16:37.668943: Pseudo dice [np.float32(0.663), np.float32(0.0), np.float32(0.0), np.float32(0.4794), np.float32(0.0), np.float32(0.0611), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0341), np.float32(0.0), np.float32(0.14), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2352), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5503), np.float32(0.0273)] 
+2026-10-16 20:16:37.669445: Epoch time: 209.26 s 
+2026-10-16 20:16:37.669873: Yayy! New best EMA pseudo Dice: 0.020899999886751175 
+2026-10-16 20:16:38.996876:  
+2026-10-16 20:16:38.997566: Epoch 21 
+2026-10-16 20:16:38.997962: Current learning rate: 0.00981 
+2026-10-16 20:20:25.724043: train_loss -0.0812 
+2026-10-16 20:20:25.724588: val_loss -0.0773 
+2026-10-16 20:20:25.725101: Pseudo dice [np.float32(0.7822), np.float32(0.0), np.float32(0.0008), np.float32(0.4428), np.float32(0.0), np.float32(0.3135), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0187), np.float32(0.0), np.float32(0.1788), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2574), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5433), np.float32(0.0421)] 
+2026-10-16 20:20:25.726084: Epoch time: 226.73 s 
+2026-10-16 20:20:25.726295: Yayy! New best EMA pseudo Dice: 0.026000000536441803 
+2026-10-16 20:20:27.247266:  
+2026-10-16 20:20:27.247908: Epoch 22 
+2026-10-16 20:20:27.248387: Current learning rate: 0.0098 
+2026-10-16 20:24:15.345286: train_loss -0.0906 
+2026-10-16 20:24:15.346362: val_loss -0.0889 
+2026-10-16 20:24:15.346906: Pseudo dice [np.float32(0.8312), np.float32(0.0), np.float32(0.0163), np.float32(0.502), np.float32(0.0), np.float32(0.6112), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1544), np.float32(0.0), np.float32(0.0044), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3673), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4712), np.float32(0.336)] 
+2026-10-16 20:24:15.347508: Epoch time: 228.1 s 
+2026-10-16 20:24:15.347954: Yayy! New best EMA pseudo Dice: 0.032499998807907104 
+2026-10-16 20:24:18.310577:  
+2026-10-16 20:24:18.311092: Epoch 23 
+2026-10-16 20:24:18.311739: Current learning rate: 0.00979 
+2026-10-16 20:28:07.306386: train_loss -0.0949 
+2026-10-16 20:28:07.307380: val_loss -0.0778 
+2026-10-16 20:28:07.307996: Pseudo dice [np.float32(0.8081), np.float32(0.0), np.float32(0.0886), np.float32(0.3423), np.float32(0.0), np.float32(0.3058), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1778), np.float32(0.0), np.float32(1e-04), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0056), np.float32(0.3534), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4027), np.float32(0.3327)] 
+2026-10-16 20:28:07.308473: Epoch time: 229.0 s 
+2026-10-16 20:28:07.308840: Yayy! New best EMA pseudo Dice: 0.03709999844431877 
+2026-10-16 20:28:08.642705:  
+2026-10-16 20:28:08.643452: Epoch 24 
+2026-10-16 20:28:08.643884: Current learning rate: 0.00978 
+2026-10-16 20:31:50.611093: train_loss -0.1025 
+2026-10-16 20:31:50.612234: val_loss -0.0799 
+2026-10-16 20:31:50.612606: Pseudo dice [np.float32(0.8387), np.float32(0.0), np.float32(0.0708), np.float32(0.3988), np.float32(0.0), np.float32(0.1651), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0016), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1855), np.float32(0.0), np.float32(0.0035), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.075), np.float32(0.4311), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4718), np.float32(0.2014)] 
+2026-10-16 20:31:50.612854: Epoch time: 221.97 s 
+2026-10-16 20:31:50.613147: Yayy! New best EMA pseudo Dice: 0.04129999876022339 
+2026-10-16 20:31:51.977209:  
+2026-10-16 20:31:51.978110: Epoch 25 
+2026-10-16 20:31:51.978532: Current learning rate: 0.00977 
+2026-10-16 20:35:23.971658: train_loss -0.107 
+2026-10-16 20:35:23.972287: val_loss -0.0927 
+2026-10-16 20:35:23.973244: Pseudo dice [np.float32(0.7751), np.float32(0.0058), np.float32(0.1176), np.float32(0.5532), np.float32(0.0), np.float32(0.384), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1088), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1128), np.float32(0.0), np.float32(0.2041), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2418), np.float32(0.2086), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3358), np.float32(0.3347)] 
+2026-10-16 20:35:23.973879: Epoch time: 212.0 s 
+2026-10-16 20:35:23.974196: Yayy! New best EMA pseudo Dice: 0.04659999907016754 
+2026-10-16 20:35:25.339798:  
+2026-10-16 20:35:25.340689: Epoch 26 
+2026-10-16 20:35:25.341105: Current learning rate: 0.00977 
+2026-10-16 20:39:15.136322: train_loss -0.1213 
+2026-10-16 20:39:15.137001: val_loss -0.0943 
+2026-10-16 20:39:15.137572: Pseudo dice [np.float32(0.7535), np.float32(0.0062), np.float32(0.2449), np.float32(0.473), np.float32(0.0), np.float32(0.3998), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3455), np.float32(0.0201), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.353), np.float32(0.0), np.float32(0.0273), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1032), np.float32(0.4831), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3981), np.float32(0.3871)] 
+2026-10-16 20:39:15.138684: Epoch time: 229.8 s 
+2026-10-16 20:39:15.139055: Yayy! New best EMA pseudo Dice: 0.05299999937415123 
+2026-10-16 20:39:18.798093:  
+2026-10-16 20:39:18.798788: Epoch 27 
+2026-10-16 20:39:18.799182: Current learning rate: 0.00976 
+2026-10-16 20:43:03.558635: train_loss -0.1282 
+2026-10-16 20:43:03.559327: val_loss -0.124 
+2026-10-16 20:43:03.560027: Pseudo dice [np.float32(0.8094), np.float32(0.2223), np.float32(0.2252), np.float32(0.5755), np.float32(0.0), np.float32(0.5694), np.float32(0.3622), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4249), np.float32(1e-04), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3589), np.float32(0.0), np.float32(0.019), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2494), np.float32(0.4372), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5816), np.float32(0.5456)] 
+2026-10-16 20:43:03.561338: Epoch time: 224.76 s 
+2026-10-16 20:43:03.561981: Yayy! New best EMA pseudo Dice: 0.06270000338554382 
+2026-10-16 20:43:05.454574:  
+2026-10-16 20:43:05.456444: Epoch 28 
+2026-10-16 20:43:05.456926: Current learning rate: 0.00975 
+2026-10-16 20:46:35.799014: train_loss -0.1517 
+2026-10-16 20:46:35.799584: val_loss -0.0926 
+2026-10-16 20:46:35.800111: Pseudo dice [np.float32(0.8222), np.float32(0.2291), np.float32(0.1846), np.float32(0.2338), np.float32(0.2762), np.float32(0.2054), np.float32(0.3937), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4237), np.float32(0.0043), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4201), np.float32(0.0), np.float32(0.0069), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2497), np.float32(0.1317), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2727), np.float32(0.1895)] 
+2026-10-16 20:46:35.800874: Epoch time: 210.35 s 
+2026-10-16 20:46:35.801299: Yayy! New best EMA pseudo Dice: 0.06759999692440033 
+2026-10-16 20:46:37.342603:  
+2026-10-16 20:46:37.343391: Epoch 29 
+2026-10-16 20:46:37.343766: Current learning rate: 0.00974 
+2026-10-16 20:50:18.084190: train_loss -0.171 
+2026-10-16 20:50:18.085325: val_loss -0.1397 
+2026-10-16 20:50:18.085919: Pseudo dice [np.float32(0.848), np.float32(0.426), np.float32(0.2915), np.float32(0.4849), np.float32(0.4811), np.float32(0.5156), np.float32(0.3352), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4534), np.float32(0.0857), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3914), np.float32(0.0044), np.float32(0.2036), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2328), np.float32(0.3899), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5286), np.float32(0.3974)] 
+2026-10-16 20:50:18.086180: Epoch time: 220.74 s 
+2026-10-16 20:50:18.086519: Yayy! New best EMA pseudo Dice: 0.07769999653100967 
+2026-10-16 20:50:19.565563:  
+2026-10-16 20:50:19.570626: Epoch 30 
+2026-10-16 20:50:19.571202: Current learning rate: 0.00973 
+2026-10-16 20:54:25.829702: train_loss -0.1795 
+2026-10-16 20:54:25.830564: val_loss -0.1219 
+2026-10-16 20:54:25.831121: Pseudo dice [np.float32(0.8223), np.float32(0.3251), np.float32(0.1474), np.float32(0.4126), np.float32(0.3714), np.float32(0.3844), np.float32(0.2511), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4293), np.float32(0.3068), np.float32(1e-04), np.float32(0.0004), np.float32(0.0), np.float32(0.0), np.float32(0.4434), np.float32(0.0385), np.float32(0.0062), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4743), np.float32(0.314), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0007), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4468), np.float32(0.2152)] 
+2026-10-16 20:54:25.831598: Epoch time: 246.27 s 
+2026-10-16 20:54:25.832184: Yayy! New best EMA pseudo Dice: 0.08489999920129776 
+2026-10-16 20:54:27.106395:  
+2026-10-16 20:54:27.107138: Epoch 31 
+2026-10-16 20:54:27.107432: Current learning rate: 0.00972 
+2026-10-16 20:58:22.271142: train_loss -0.193 
+2026-10-16 20:58:22.272058: val_loss -0.1902 
+2026-10-16 20:58:22.273195: Pseudo dice [np.float32(0.8181), np.float32(0.4559), np.float32(0.4603), np.float32(0.7023), np.float32(0.5067), np.float32(0.6417), np.float32(0.4904), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5317), np.float32(0.4733), np.float32(0.2653), np.float32(0.245), np.float32(0.0), np.float32(0.0), np.float32(0.3886), np.float32(0.2091), np.float32(0.4164), np.float32(0.0), np.float32(0.036), np.float32(0.0), np.float32(0.2288), np.float32(0.5517), np.float32(0.0), np.float32(0.0054), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6373), np.float32(0.6557)] 
+2026-10-16 20:58:22.273796: Epoch time: 235.17 s 
+2026-10-16 20:58:22.274146: Yayy! New best EMA pseudo Dice: 0.1005999967455864 
+2026-10-16 20:58:25.023417:  
+2026-10-16 20:58:25.024555: Epoch 32 
+2026-10-16 20:58:25.024901: Current learning rate: 0.00971 
+2026-10-16 21:02:33.395450: train_loss -0.2048 
+2026-10-16 21:02:33.396512: val_loss -0.21 
+2026-10-16 21:02:33.397053: Pseudo dice [np.float32(0.8685), np.float32(0.5936), np.float32(0.5233), np.float32(0.7595), np.float32(0.6593), np.float32(0.7257), np.float32(0.6391), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4858), np.float32(0.5309), np.float32(0.343), np.float32(0.0217), np.float32(0.0), np.float32(0.0), np.float32(0.5012), np.float32(0.2002), np.float32(0.3035), np.float32(0.0), np.float32(0.0116), np.float32(0.0202), np.float32(0.346), np.float32(0.3868), np.float32(0.0), np.float32(0.0003), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0063), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6117), np.float32(0.6319)] 
+2026-10-16 21:02:33.397497: Epoch time: 248.37 s 
+2026-10-16 21:02:33.397801: Yayy! New best EMA pseudo Dice: 0.1160999983549118 
+2026-10-16 21:02:34.717839:  
+2026-10-16 21:02:34.718584: Epoch 33 
+2026-10-16 21:02:34.718941: Current learning rate: 0.0097 
+2026-10-16 21:06:47.459382: train_loss -0.2221 
+2026-10-16 21:06:47.460287: val_loss -0.1661 
+2026-10-16 21:06:47.460667: Pseudo dice [np.float32(0.8415), np.float32(0.3831), np.float32(0.3888), np.float32(0.5312), np.float32(0.4683), np.float32(0.3349), np.float32(0.3783), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4933), np.float32(0.1389), np.float32(0.365), np.float32(0.0004), np.float32(0.0), np.float32(0.0), np.float32(0.4783), np.float32(0.2772), np.float32(0.4102), np.float32(0.0389), np.float32(0.1072), np.float32(0.1365), np.float32(0.125), np.float32(0.4258), np.float32(0.0002), np.float32(0.0377), np.float32(0.0), np.float32(0.0), np.float32(0.0501), np.float32(0.1132), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5128), np.float32(0.3715)] 
+2026-10-16 21:06:47.461071: Epoch time: 252.74 s 
+2026-10-16 21:06:47.461256: Yayy! New best EMA pseudo Dice: 0.125 
+2026-10-16 21:06:48.872058:  
+2026-10-16 21:06:48.872733: Epoch 34 
+2026-10-16 21:06:48.873096: Current learning rate: 0.00969 
+2026-10-16 21:10:44.429559: train_loss -0.2361 
+2026-10-16 21:10:44.430439: val_loss -0.2315 
+2026-10-16 21:10:44.430990: Pseudo dice [np.float32(0.784), np.float32(0.6465), np.float32(0.5503), np.float32(0.7563), np.float32(0.5938), np.float32(0.7527), np.float32(0.6753), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5334), np.float32(0.3639), np.float32(0.4081), np.float32(0.001), np.float32(0.0), np.float32(0.0), np.float32(0.6059), np.float32(0.3223), np.float32(0.585), np.float32(0.1824), np.float32(0.1143), np.float32(0.1534), np.float32(0.085), np.float32(0.4839), np.float32(0.0911), np.float32(0.1566), np.float32(0.0), np.float32(0.0), np.float32(0.0502), np.float32(0.1979), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7688), np.float32(0.7503)] 
+2026-10-16 21:10:44.431571: Epoch time: 235.56 s 
+2026-10-16 21:10:44.431998: Yayy! New best EMA pseudo Dice: 0.1420000046491623 
+2026-10-16 21:10:45.760820:  
+2026-10-16 21:10:45.761514: Epoch 35 
+2026-10-16 21:10:45.761876: Current learning rate: 0.00968 
+2026-10-16 21:14:23.031702: train_loss -0.2342 
+2026-10-16 21:14:23.032632: val_loss -0.1754 
+2026-10-16 21:14:23.033679: Pseudo dice [np.float32(0.8167), np.float32(0.455), np.float32(0.3572), np.float32(0.4819), np.float32(0.3931), np.float32(0.3682), np.float32(0.4202), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4284), np.float32(0.4492), np.float32(0.432), np.float32(0.0395), np.float32(0.0), np.float32(0.0), np.float32(0.5217), np.float32(0.3017), np.float32(0.4186), np.float32(0.1495), np.float32(0.2049), np.float32(0.0032), np.float32(0.2349), np.float32(0.4043), np.float32(0.1886), np.float32(0.1125), np.float32(0.0), np.float32(0.0), np.float32(0.0224), np.float32(0.253), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3559), np.float32(0.3849)] 
+2026-10-16 21:14:23.034168: Epoch time: 217.27 s 
+2026-10-16 21:14:23.034486: Yayy! New best EMA pseudo Dice: 0.15060000121593475 
+2026-10-16 21:14:26.379578:  
+2026-10-16 21:14:26.380620: Epoch 36 
+2026-10-16 21:14:26.381124: Current learning rate: 0.00968 
+2026-10-16 21:18:04.422174: train_loss -0.2479 
+2026-10-16 21:18:04.422946: val_loss -0.2377 
+2026-10-16 21:18:04.423548: Pseudo dice [np.float32(0.8604), np.float32(0.5917), np.float32(0.5441), np.float32(0.7306), np.float32(0.57), np.float32(0.7017), np.float32(0.5936), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5425), np.float32(0.5059), np.float32(0.3769), np.float32(0.1891), np.float32(0.0), np.float32(0.0), np.float32(0.5611), np.float32(0.2608), np.float32(0.5512), np.float32(0.2609), np.float32(0.1629), np.float32(0.2672), np.float32(0.274), np.float32(0.3542), np.float32(0.1619), np.float32(0.1726), np.float32(0.0), np.float32(0.0), np.float32(0.0997), np.float32(0.216), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7296), np.float32(0.7203)] 
+2026-10-16 21:18:04.424268: Epoch time: 218.04 s 
+2026-10-16 21:18:04.424744: Yayy! New best EMA pseudo Dice: 0.16609999537467957 
+2026-10-16 21:18:05.995376:  
+2026-10-16 21:18:05.996251: Epoch 37 
+2026-10-16 21:18:05.996586: Current learning rate: 0.00967 
+2026-10-16 21:21:43.426804: train_loss -0.2791 
+2026-10-16 21:21:43.427663: val_loss -0.1726 
+2026-10-16 21:21:43.428301: Pseudo dice [np.float32(0.8432), np.float32(0.4447), np.float32(0.3563), np.float32(0.4479), np.float32(0.3261), np.float32(0.4443), np.float32(0.2221), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.501), np.float32(0.3531), np.float32(0.3914), np.float32(0.2177), np.float32(0.0), np.float32(0.0), np.float32(0.3866), np.float32(0.3166), np.float32(0.3171), np.float32(0.2898), np.float32(0.1643), np.float32(0.2669), np.float32(0.2756), np.float32(0.3097), np.float32(0.2813), np.float32(0.2122), np.float32(0.0), np.float32(0.0), np.float32(0.0549), np.float32(0.1862), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.407), np.float32(0.3651)] 
+2026-10-16 21:21:43.428661: Epoch time: 217.43 s 
+2026-10-16 21:21:43.429091: Yayy! New best EMA pseudo Dice: 0.17270000278949738 
+2026-10-16 21:21:44.815874:  
+2026-10-16 21:21:44.816479: Epoch 38 
+2026-10-16 21:21:44.816863: Current learning rate: 0.00966 
+2026-10-16 21:25:30.171652: train_loss -0.2543 
+2026-10-16 21:25:30.172274: val_loss -0.0721 
+2026-10-16 21:25:30.173342: Pseudo dice [np.float32(0.7905), np.float32(0.094), np.float32(0.1543), np.float32(0.1722), np.float32(0.0207), np.float32(0.0938), np.float32(0.0823), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2896), np.float32(0.093), np.float32(0.375), np.float32(0.0929), np.float32(0.0), np.float32(0.0), np.float32(0.0197), np.float32(0.1235), np.float32(0.1086), np.float32(0.2081), np.float32(0.0847), np.float32(0.1505), np.float32(0.2511), np.float32(0.2488), np.float32(0.0347), np.float32(0.044), np.float32(0.0), np.float32(0.0), np.float32(0.1096), np.float32(0.1663), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0443), np.float32(0.1216)] 
+2026-10-16 21:25:30.173673: Epoch time: 225.36 s 
+2026-10-16 21:25:31.021023:  
+2026-10-16 21:25:31.021760: Epoch 39 
+2026-10-16 21:25:31.022138: Current learning rate: 0.00965 
+2026-10-16 21:29:30.235684: train_loss -0.2587 
+2026-10-16 21:29:30.236404: val_loss -0.1669 
+2026-10-16 21:29:30.236854: Pseudo dice [np.float32(0.803), np.float32(0.3626), np.float32(0.3199), np.float32(0.3822), np.float32(0.2984), np.float32(0.3484), np.float32(0.324), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3792), np.float32(0.4306), np.float32(0.3721), np.float32(0.1193), np.float32(0.0), np.float32(0.0), np.float32(0.2833), np.float32(0.2724), np.float32(0.3137), np.float32(0.3097), np.float32(0.1481), np.float32(0.2046), np.float32(0.2131), np.float32(0.5236), np.float32(0.2674), np.float32(0.2357), np.float32(0.0), np.float32(0.0), np.float32(0.0408), np.float32(0.2723), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3836), np.float32(0.4368)] 
+2026-10-16 21:29:30.237313: Epoch time: 239.22 s 
+2026-10-16 21:29:31.100128:  
+2026-10-16 21:29:31.100804: Epoch 40 
+2026-10-16 21:29:31.101138: Current learning rate: 0.00964 
+2026-10-16 21:33:40.478375: train_loss -0.2845 
+2026-10-16 21:33:40.479313: val_loss -0.2843 
+2026-10-16 21:33:40.480011: Pseudo dice [np.float32(0.8366), np.float32(0.706), np.float32(0.6101), np.float32(0.8042), np.float32(0.6792), np.float32(0.7904), np.float32(0.6909), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6079), np.float32(0.6266), np.float32(0.4177), np.float32(0.2337), np.float32(0.0), np.float32(0.0), np.float32(0.6736), np.float32(0.3516), np.float32(0.6422), np.float32(0.4119), np.float32(0.4411), np.float32(0.45), np.float32(0.3237), np.float32(0.5117), np.float32(0.4154), np.float32(0.3293), np.float32(0.0), np.float32(0.0), np.float32(0.2877), np.float32(0.2924), np.float32(0.0), np.float32(0.0), np.float32(0.0244), np.float32(0.0), np.float32(0.8062), np.float32(0.7532)] 
+2026-10-16 21:33:40.480443: Epoch time: 249.38 s 
+2026-10-16 21:33:40.480828: Yayy! New best EMA pseudo Dice: 0.1931000053882599 
+2026-10-16 21:33:41.990636:  
+2026-10-16 21:33:41.991446: Epoch 41 
+2026-10-16 21:33:41.993737: Current learning rate: 0.00963 
+2026-10-16 21:37:22.024080: train_loss -0.2563 
+2026-10-16 21:37:22.024637: val_loss -0.2415 
+2026-10-16 21:37:22.025172: Pseudo dice [np.float32(0.8516), np.float32(0.6271), np.float32(0.5673), np.float32(0.7262), np.float32(0.5887), np.float32(0.7389), np.float32(0.6077), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5493), np.float32(0.5869), np.float32(0.4527), np.float32(0.2781), np.float32(0.0), np.float32(0.0), np.float32(0.5411), np.float32(0.2647), np.float32(0.5505), np.float32(0.276), np.float32(0.3096), np.float32(0.3704), np.float32(0.4812), np.float32(0.4665), np.float32(0.315), np.float32(0.3462), np.float32(0.0), np.float32(0.0), np.float32(0.1169), np.float32(0.2864), np.float32(0.0), np.float32(0.0), np.float32(0.0545), np.float32(0.0), np.float32(0.6237), np.float32(0.4896)] 
+2026-10-16 21:37:22.025786: Epoch time: 220.04 s 
+2026-10-16 21:37:22.026110: Yayy! New best EMA pseudo Dice: 0.20730000734329224 
+2026-10-16 21:37:23.374144:  
+2026-10-16 21:37:23.374786: Epoch 42 
+2026-10-16 21:37:23.375083: Current learning rate: 0.00962 
+2026-10-16 21:41:05.411139: train_loss -0.3094 
+2026-10-16 21:41:05.412212: val_loss -0.2934 
+2026-10-16 21:41:05.412705: Pseudo dice [np.float32(0.841), np.float32(0.6704), np.float32(0.647), np.float32(0.7752), np.float32(0.6646), np.float32(0.7639), np.float32(0.6988), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6232), np.float32(0.5842), np.float32(0.5368), np.float32(0.3856), np.float32(0.0), np.float32(0.0), np.float32(0.6642), np.float32(0.3642), np.float32(0.6806), np.float32(0.3378), np.float32(0.3493), np.float32(0.437), np.float32(0.6302), np.float32(0.6889), np.float32(0.3634), np.float32(0.3527), np.float32(0.0), np.float32(0.0), np.float32(0.157), np.float32(0.4448), np.float32(0.0), np.float32(0.0), np.float32(0.3105), np.float32(0.0045), np.float32(0.7555), np.float32(0.7309)] 
+2026-10-16 21:41:05.413024: Epoch time: 222.04 s 
+2026-10-16 21:41:05.413455: Yayy! New best EMA pseudo Dice: 0.22669999301433563 
+2026-10-16 21:41:06.786945:  
+2026-10-16 21:41:06.787803: Epoch 43 
+2026-10-16 21:41:06.788137: Current learning rate: 0.00961 
+2026-10-16 21:44:52.789324: train_loss -0.3384 
+2026-10-16 21:44:52.790358: val_loss -0.3162 
+2026-10-16 21:44:52.790958: Pseudo dice [np.float32(0.8544), np.float32(0.6587), np.float32(0.6144), np.float32(0.7156), np.float32(0.6193), np.float32(0.6491), np.float32(0.5814), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6066), np.float32(0.6201), np.float32(0.5082), np.float32(0.4171), np.float32(0.0), np.float32(0.0), np.float32(0.5977), np.float32(0.3338), np.float32(0.5532), np.float32(0.3031), np.float32(0.3023), np.float32(0.3513), np.float32(0.6484), np.float32(0.7006), np.float32(0.4266), np.float32(0.3374), np.float32(0.0), np.float32(0.0), np.float32(0.2216), np.float32(0.4409), np.float32(0.0), np.float32(0.0), np.float32(0.1334), np.float32(0.2236), np.float32(0.7434), np.float32(0.693)] 
+2026-10-16 21:44:52.791509: Epoch time: 226.0 s 
+2026-10-16 21:44:52.791795: Yayy! New best EMA pseudo Dice: 0.2425999939441681 
+2026-10-16 21:44:54.586104:  
+2026-10-16 21:44:54.586871: Epoch 44 
+2026-10-16 21:44:54.587283: Current learning rate: 0.0096 
+2026-10-16 21:48:34.608715: train_loss -0.3767 
+2026-10-16 21:48:34.609310: val_loss -0.373 
+2026-10-16 21:48:34.609804: Pseudo dice [np.float32(0.8211), np.float32(0.721), np.float32(0.6468), np.float32(0.8364), np.float32(0.6911), np.float32(0.8242), np.float32(0.7522), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6578), np.float32(0.6632), np.float32(0.5113), np.float32(0.5029), np.float32(0.0), np.float32(0.0), np.float32(0.6706), np.float32(0.4113), np.float32(0.6704), np.float32(0.4225), np.float32(0.3152), np.float32(0.4622), np.float32(0.5985), np.float32(0.5986), np.float32(0.42), np.float32(0.4678), np.float32(0.0), np.float32(0.0), np.float32(0.2729), np.float32(0.3593), np.float32(0.0), np.float32(0.0), np.float32(0.5244), np.float32(0.5221), np.float32(0.8235), np.float32(0.7861)] 
+2026-10-16 21:48:34.610074: Epoch time: 220.02 s 
+2026-10-16 21:48:34.610761: Yayy! New best EMA pseudo Dice: 0.26260000467300415 
+2026-10-16 21:48:35.906902:  
+2026-10-16 21:48:35.909074: Epoch 45 
+2026-10-16 21:48:35.911066: Current learning rate: 0.00959 
+2026-10-16 21:52:07.047064: train_loss -0.4387 
+2026-10-16 21:52:07.047973: val_loss -0.3765 
+2026-10-16 21:52:07.048581: Pseudo dice [np.float32(0.8311), np.float32(0.4865), np.float32(0.4925), np.float32(0.6268), np.float32(0.5258), np.float32(0.5853), np.float32(0.5305), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5148), np.float32(0.5051), np.float32(0.5003), np.float32(0.4922), np.float32(0.0), np.float32(0.0), np.float32(0.5682), np.float32(0.2755), np.float32(0.5369), np.float32(0.3449), np.float32(0.277), np.float32(0.3079), np.float32(0.3417), np.float32(0.5214), np.float32(0.3528), np.float32(0.2773), np.float32(0.0), np.float32(0.0), np.float32(0.2053), np.float32(0.3875), np.float32(0.0), np.float32(0.0), np.float32(0.3864), np.float32(0.2962), np.float32(0.5931), np.float32(0.5811)] 
+2026-10-16 21:52:07.049141: Epoch time: 211.14 s 
+2026-10-16 21:52:07.049535: Yayy! New best EMA pseudo Dice: 0.27059999108314514 
+2026-10-16 21:52:10.365945:  
+2026-10-16 21:52:10.366601: Epoch 46 
+2026-10-16 21:52:10.366992: Current learning rate: 0.00959 
+2026-10-16 21:55:38.176710: train_loss -0.4712 
+2026-10-16 21:55:38.177835: val_loss -0.2462 
+2026-10-16 21:55:38.178592: Pseudo dice [np.float32(0.8723), np.float32(0.1928), np.float32(0.2469), np.float32(0.3726), np.float32(0.2265), np.float32(0.3548), np.float32(0.1671), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3872), np.float32(0.3571), np.float32(0.3212), np.float32(0.3965), np.float32(0.0), np.float32(0.0), np.float32(0.2266), np.float32(0.1387), np.float32(0.1224), np.float32(0.1678), np.float32(0.0748), np.float32(0.0966), np.float32(0.3221), np.float32(0.2339), np.float32(0.2562), np.float32(0.133), np.float32(0.0), np.float32(0.0), np.float32(0.0408), np.float32(0.0955), np.float32(0.0), np.float32(0.0), np.float32(0.2511), np.float32(0.1669), np.float32(0.4024), np.float32(0.34)] 
+2026-10-16 21:55:38.179017: Epoch time: 207.81 s 
+2026-10-16 21:55:39.015979:  
+2026-10-16 21:55:39.016807: Epoch 47 
+2026-10-16 21:55:39.017191: Current learning rate: 0.00958 
+2026-10-16 21:59:16.921276: train_loss -0.4885 
+2026-10-16 21:59:16.922248: val_loss -0.3713 
+2026-10-16 21:59:16.922919: Pseudo dice [np.float32(0.8338), np.float32(0.4832), np.float32(0.4813), np.float32(0.5927), np.float32(0.4766), np.float32(0.539), np.float32(0.4416), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5016), np.float32(0.4634), np.float32(0.4745), np.float32(0.4391), np.float32(0.0), np.float32(0.0), np.float32(0.5247), np.float32(0.3729), np.float32(0.4232), np.float32(0.3506), np.float32(0.38), np.float32(0.3432), np.float32(0.3646), np.float32(0.4933), np.float32(0.4117), np.float32(0.3404), np.float32(0.0), np.float32(0.0), np.float32(0.2023), np.float32(0.3954), np.float32(0.0), np.float32(0.0), np.float32(0.4134), np.float32(0.2969), np.float32(0.5659), np.float32(0.5455)] 
+2026-10-16 21:59:16.923432: Epoch time: 217.91 s 
+2026-10-16 21:59:17.745006:  
+2026-10-16 21:59:17.745636: Epoch 48 
+2026-10-16 21:59:17.746064: Current learning rate: 0.00957 
+2026-10-16 22:03:11.261516: train_loss -0.4943 
+2026-10-16 22:03:11.262283: val_loss -0.4413 
+2026-10-16 22:03:11.263015: Pseudo dice [np.float32(0.8725), np.float32(0.6219), np.float32(0.6091), np.float32(0.7658), np.float32(0.62), np.float32(0.7146), np.float32(0.6783), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6058), np.float32(0.5862), np.float32(0.4689), np.float32(0.4628), np.float32(0.0), np.float32(0.0), np.float32(0.5775), np.float32(0.4366), np.float32(0.5465), np.float32(0.4165), np.float32(0.4738), np.float32(0.4592), np.float32(0.473), np.float32(0.4844), np.float32(0.3477), np.float32(0.3756), np.float32(0.0), np.float32(0.0), np.float32(0.3952), np.float32(0.5125), np.float32(0.0), np.float32(0.0), np.float32(0.544), np.float32(0.4604), np.float32(0.7079), np.float32(0.638)] 
+2026-10-16 22:03:11.263606: Epoch time: 233.52 s 
+2026-10-16 22:03:11.264131: Yayy! New best EMA pseudo Dice: 0.28459998965263367 
+2026-10-16 22:03:12.918149:  
+2026-10-16 22:03:12.918779: Epoch 49 
+2026-10-16 22:03:12.919245: Current learning rate: 0.00956 
+2026-10-16 22:06:54.696235: train_loss -0.4894 
+2026-10-16 22:06:54.696963: val_loss -0.3492 
+2026-10-16 22:06:54.697472: Pseudo dice [np.float32(0.8546), np.float32(0.421), np.float32(0.4555), np.float32(0.6015), np.float32(0.424), np.float32(0.5401), np.float32(0.4477), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5448), np.float32(0.4663), np.float32(0.4737), np.float32(0.4056), np.float32(0.0), np.float32(0.0), np.float32(0.4909), np.float32(0.2741), np.float32(0.4676), np.float32(0.2609), np.float32(0.414), np.float32(0.3878), np.float32(0.277), np.float32(0.3489), np.float32(0.3729), np.float32(0.2747), np.float32(0.0), np.float32(0.0), np.float32(0.3253), np.float32(0.4152), np.float32(0.0), np.float32(0.0), np.float32(0.3966), np.float32(0.3983), np.float32(0.4036), np.float32(0.3961)] 
+2026-10-16 22:06:54.698035: Epoch time: 221.78 s 
+2026-10-16 22:06:55.227288: Yayy! New best EMA pseudo Dice: 0.2881999909877777 
+2026-10-16 22:06:56.497087:  
+2026-10-16 22:06:56.498163: Epoch 50 
+2026-10-16 22:06:56.498676: Current learning rate: 0.00955 
+2026-10-16 22:10:49.027158: train_loss -0.4894 
+2026-10-16 22:10:49.028013: val_loss -0.4888 
+2026-10-16 22:10:49.028529: Pseudo dice [np.float32(0.8666), np.float32(0.7317), np.float32(0.6598), np.float32(0.8322), np.float32(0.7593), np.float32(0.8193), np.float32(0.7884), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6657), np.float32(0.6714), np.float32(0.5631), np.float32(0.618), np.float32(0.0), np.float32(0.0), np.float32(0.7116), np.float32(0.5141), np.float32(0.6907), np.float32(0.512), np.float32(0.4934), np.float32(0.5276), np.float32(0.6937), np.float32(0.7061), np.float32(0.4539), np.float32(0.422), np.float32(0.0), np.float32(0.0), np.float32(0.2683), np.float32(0.4468), np.float32(0.0), np.float32(0.0), np.float32(0.6052), np.float32(0.5402), np.float32(0.8162), np.float32(0.7911)] 
+2026-10-16 22:10:49.028898: Epoch time: 232.53 s 
+2026-10-16 22:10:49.029360: Yayy! New best EMA pseudo Dice: 0.30709999799728394 
+2026-10-16 22:10:50.345498:  
+2026-10-16 22:10:50.346363: Epoch 51 
+2026-10-16 22:10:50.346747: Current learning rate: 0.00954 
+2026-10-16 22:15:00.572397: train_loss -0.4632 
+2026-10-16 22:15:00.573654: val_loss -0.4606 
+2026-10-16 22:15:00.574161: Pseudo dice [np.float32(0.8639), np.float32(0.6737), np.float32(0.6134), np.float32(0.7699), np.float32(0.6175), np.float32(0.7727), np.float32(0.6779), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6089), np.float32(0.5838), np.float32(0.4943), np.float32(0.5085), np.float32(0.0), np.float32(0.0), np.float32(0.6677), np.float32(0.5158), np.float32(0.6192), np.float32(0.5174), np.float32(0.5122), np.float32(0.4788), np.float32(0.6729), np.float32(0.6826), np.float32(0.4317), np.float32(0.4512), np.float32(0.0), np.float32(0.0), np.float32(0.4337), np.float32(0.5117), np.float32(0.0), np.float32(0.0), np.float32(0.5318), np.float32(0.5241), np.float32(0.7923), np.float32(0.7605)] 
+2026-10-16 22:15:00.574572: Epoch time: 250.23 s 
+2026-10-16 22:15:00.574858: Yayy! New best EMA pseudo Dice: 0.3215999901294708 
+2026-10-16 22:15:01.902752:  
+2026-10-16 22:15:01.903308: Epoch 52 
+2026-10-16 22:15:01.903921: Current learning rate: 0.00953 
+2026-10-16 22:18:45.812335: train_loss -0.4886 
+2026-10-16 22:18:45.813113: val_loss -0.3366 
+2026-10-16 22:18:45.813878: Pseudo dice [np.float32(0.8484), np.float32(0.3862), np.float32(0.3943), np.float32(0.5545), np.float32(0.4025), np.float32(0.5089), np.float32(0.3905), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5112), np.float32(0.4387), np.float32(0.5128), np.float32(0.4588), np.float32(0.0), np.float32(0.0), np.float32(0.421), np.float32(0.2686), np.float32(0.4005), np.float32(0.2996), np.float32(0.2645), np.float32(0.3445), np.float32(0.3097), np.float32(0.3223), np.float32(0.3827), np.float32(0.2771), np.float32(0.0), np.float32(0.0), np.float32(0.2911), np.float32(0.4448), np.float32(0.0), np.float32(0.0), np.float32(0.3874), np.float32(0.2021), np.float32(0.4509), np.float32(0.4425)] 
+2026-10-16 22:18:45.814415: Epoch time: 223.91 s 
+2026-10-16 22:18:46.662760:  
+2026-10-16 22:18:46.663274: Epoch 53 
+2026-10-16 22:18:46.663765: Current learning rate: 0.00952 
+2026-10-16 22:22:49.334303: train_loss -0.5068 
+2026-10-16 22:22:49.334884: val_loss -0.4682 
+2026-10-16 22:22:49.335356: Pseudo dice [np.float32(0.8667), np.float32(0.6768), np.float32(0.6172), np.float32(0.8145), np.float32(0.6528), np.float32(0.7984), np.float32(0.7063), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6885), np.float32(0.6853), np.float32(0.5935), np.float32(0.6145), np.float32(0.0), np.float32(0.0), np.float32(0.6651), np.float32(0.4821), np.float32(0.6533), np.float32(0.5178), np.float32(0.5377), np.float32(0.5213), np.float32(0.5381), np.float32(0.5885), np.float32(0.4688), np.float32(0.3865), np.float32(0.0), np.float32(0.0), np.float32(0.4846), np.float32(0.3641), np.float32(0.0), np.float32(0.0), np.float32(0.6548), np.float32(0.5928), np.float32(0.7896), np.float32(0.7739)] 
+2026-10-16 22:22:49.336188: Epoch time: 242.67 s 
+2026-10-16 22:22:49.336378: Yayy! New best EMA pseudo Dice: 0.3343000113964081 
+2026-10-16 22:22:51.640958:  
+2026-10-16 22:22:51.641634: Epoch 54 
+2026-10-16 22:22:51.642108: Current learning rate: 0.00951 
+2026-10-16 22:26:49.597436: train_loss -0.4968 
+2026-10-16 22:26:49.598123: val_loss -0.3001 
+2026-10-16 22:26:49.598945: Pseudo dice [np.float32(0.8444), np.float32(0.2963), np.float32(0.3032), np.float32(0.3886), np.float32(0.1922), np.float32(0.2967), np.float32(0.2412), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3467), np.float32(0.2612), np.float32(0.3246), np.float32(0.2506), np.float32(0.0), np.float32(0.0), np.float32(0.3851), np.float32(0.2873), np.float32(0.305), np.float32(0.2945), np.float32(0.1573), np.float32(0.2705), np.float32(0.2851), np.float32(0.445), np.float32(0.2978), np.float32(0.3366), np.float32(0.0), np.float32(0.0), np.float32(0.0536), np.float32(0.4679), np.float32(0.0), np.float32(0.0), np.float32(0.2021), np.float32(0.0699), np.float32(0.3027), np.float32(0.2826)] 
+2026-10-16 22:26:49.599387: Epoch time: 237.96 s 
+2026-10-16 22:26:50.436177:  
+2026-10-16 22:26:50.436855: Epoch 55 
+2026-10-16 22:26:50.437325: Current learning rate: 0.0095 
+2026-10-16 22:30:53.607364: train_loss -0.5245 
+2026-10-16 22:30:53.608217: val_loss -0.3044 
+2026-10-16 22:30:53.608720: Pseudo dice [np.float32(0.7975), np.float32(0.3688), np.float32(0.3655), np.float32(0.4028), np.float32(0.3268), np.float32(0.3661), np.float32(0.2892), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3636), np.float32(0.3152), np.float32(0.3657), np.float32(0.0838), np.float32(0.0), np.float32(0.0), np.float32(0.3938), np.float32(0.297), np.float32(0.2219), np.float32(0.0905), np.float32(0.2093), np.float32(0.2606), np.float32(0.2342), np.float32(0.5794), np.float32(0.3761), np.float32(0.3145), np.float32(0.0), np.float32(0.0), np.float32(0.0321), np.float32(0.4821), np.float32(0.0), np.float32(0.0), np.float32(0.2894), np.float32(0.1742), np.float32(0.2297), np.float32(0.397)] 
+2026-10-16 22:30:53.609104: Epoch time: 243.17 s 
+2026-10-16 22:30:54.414055:  
+2026-10-16 22:30:54.414725: Epoch 56 
+2026-10-16 22:30:54.414955: Current learning rate: 0.00949 
+2026-10-16 22:34:55.947366: train_loss -0.5132 
+2026-10-16 22:34:55.947933: val_loss -0.3745 
+2026-10-16 22:34:55.948447: Pseudo dice [np.float32(0.8578), np.float32(0.4848), np.float32(0.5174), np.float32(0.5682), np.float32(0.4403), np.float32(0.5277), np.float32(0.5083), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4667), np.float32(0.4269), np.float32(0.4271), np.float32(0.2775), np.float32(0.0), np.float32(0.0), np.float32(0.4969), np.float32(0.3947), np.float32(0.4738), np.float32(0.3905), np.float32(0.3869), np.float32(0.4171), np.float32(0.377), np.float32(0.5056), np.float32(0.3111), np.float32(0.3719), np.float32(0.0), np.float32(0.0), np.float32(0.2635), np.float32(0.5148), np.float32(0.0), np.float32(0.0), np.float32(0.3585), np.float32(0.3269), np.float32(0.5428), np.float32(0.527)] 
+2026-10-16 22:34:55.949420: Epoch time: 241.54 s 
+2026-10-16 22:34:56.764556:  
+2026-10-16 22:34:56.765241: Epoch 57 
+2026-10-16 22:34:56.765593: Current learning rate: 0.00949 
+2026-10-16 22:38:45.232285: train_loss -0.51 
+2026-10-16 22:38:45.233325: val_loss -0.4649 
+2026-10-16 22:38:45.233943: Pseudo dice [np.float32(0.8365), np.float32(0.6805), np.float32(0.6404), np.float32(0.8014), np.float32(0.6295), np.float32(0.7736), np.float32(0.7035), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6187), np.float32(0.619), np.float32(0.6005), np.float32(0.5957), np.float32(0.0), np.float32(0.0), np.float32(0.6582), np.float32(0.5177), np.float32(0.6615), np.float32(0.5094), np.float32(0.424), np.float32(0.5286), np.float32(0.6102), np.float32(0.6957), np.float32(0.4981), np.float32(0.4211), np.float32(0.0), np.float32(0.0), np.float32(0.4976), np.float32(0.48), np.float32(0.0), np.float32(0.0), np.float32(0.5272), np.float32(0.5664), np.float32(0.764), np.float32(0.7642)] 
+2026-10-16 22:38:45.234500: Epoch time: 228.47 s 
+2026-10-16 22:38:46.042085:  
+2026-10-16 22:38:46.042806: Epoch 58 
+2026-10-16 22:38:46.043233: Current learning rate: 0.00948 
+2026-10-16 22:42:31.774789: train_loss -0.54 
+2026-10-16 22:42:31.775320: val_loss -0.513 
+2026-10-16 22:42:31.775927: Pseudo dice [np.float32(0.8508), np.float32(0.7439), np.float32(0.6689), np.float32(0.8617), np.float32(0.7076), np.float32(0.8501), np.float32(0.7707), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7077), np.float32(0.7013), np.float32(0.6265), np.float32(0.6191), np.float32(0.0), np.float32(0.0), np.float32(0.7291), np.float32(0.5356), np.float32(0.6837), np.float32(0.5106), np.float32(0.5641), np.float32(0.5383), np.float32(0.71), np.float32(0.656), np.float32(0.564), np.float32(0.5551), np.float32(0.0), np.float32(0.0), np.float32(0.3207), np.float32(0.5165), np.float32(0.0), np.float32(0.0), np.float32(0.678), np.float32(0.6238), np.float32(0.8454), np.float32(0.8327)] 
+2026-10-16 22:42:31.776658: Epoch time: 225.74 s 
+2026-10-16 22:42:31.776903: Yayy! New best EMA pseudo Dice: 0.34860000014305115 
+2026-10-16 22:42:37.242758:  
+2026-10-16 22:42:37.243497: Epoch 59 
+2026-10-16 22:42:37.243869: Current learning rate: 0.00947 
+2026-10-16 22:46:16.454777: train_loss -0.5519 
+2026-10-16 22:46:16.455506: val_loss -0.3649 
+2026-10-16 22:46:16.456371: Pseudo dice [np.float32(0.8558), np.float32(0.4388), np.float32(0.4345), np.float32(0.571), np.float32(0.3735), np.float32(0.5351), np.float32(0.3855), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5164), np.float32(0.4675), np.float32(0.5018), np.float32(0.5063), np.float32(0.0), np.float32(0.0), np.float32(0.4652), np.float32(0.3616), np.float32(0.4016), np.float32(0.3338), np.float32(0.1866), np.float32(0.3961), np.float32(0.3466), np.float32(0.4438), np.float32(0.3886), np.float32(0.302), np.float32(0.0), np.float32(0.0), np.float32(0.3668), np.float32(0.4342), np.float32(0.0), np.float32(0.0), np.float32(0.4563), np.float32(0.4256), np.float32(0.5049), np.float32(0.4654)] 
+2026-10-16 22:46:16.456749: Epoch time: 219.21 s 
+2026-10-16 22:46:17.324155:  
+2026-10-16 22:46:17.324794: Epoch 60 
+2026-10-16 22:46:17.325163: Current learning rate: 0.00946 
+2026-10-16 22:49:57.958046: train_loss -0.5631 
+2026-10-16 22:49:57.958715: val_loss -0.5262 
+2026-10-16 22:49:57.959224: Pseudo dice [np.float32(0.8548), np.float32(0.7249), np.float32(0.6735), np.float32(0.8629), np.float32(0.7138), np.float32(0.8594), np.float32(0.7675), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7246), np.float32(0.6944), np.float32(0.6657), np.float32(0.6796), np.float32(0.0), np.float32(0.0), np.float32(0.7049), np.float32(0.5361), np.float32(0.6708), np.float32(0.5159), np.float32(0.5295), np.float32(0.5473), np.float32(0.6506), np.float32(0.6257), np.float32(0.4747), np.float32(0.3947), np.float32(0.0), np.float32(0.0), np.float32(0.4619), np.float32(0.4382), np.float32(0.0), np.float32(0.0), np.float32(0.661), np.float32(0.6353), np.float32(0.842), np.float32(0.8356)] 
+2026-10-16 22:49:57.960202: Epoch time: 220.64 s 
+2026-10-16 22:49:57.960386: Yayy! New best EMA pseudo Dice: 0.361299991607666 
+2026-10-16 22:49:59.956850:  
+2026-10-16 22:49:59.957528: Epoch 61 
+2026-10-16 22:49:59.957968: Current learning rate: 0.00945 
+2026-10-16 22:53:48.079370: train_loss -0.5808 
+2026-10-16 22:53:48.080192: val_loss -0.5061 
+2026-10-16 22:53:48.080767: Pseudo dice [np.float32(0.8646), np.float32(0.6993), np.float32(0.6638), np.float32(0.8034), np.float32(0.6339), np.float32(0.7801), np.float32(0.6634), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6763), np.float32(0.6424), np.float32(0.6076), np.float32(0.6037), np.float32(0.0), np.float32(0.0), np.float32(0.6603), np.float32(0.5357), np.float32(0.6425), np.float32(0.5327), np.float32(0.5316), np.float32(0.5576), np.float32(0.6713), np.float32(0.687), np.float32(0.418), np.float32(0.4578), np.float32(0.0), np.float32(0.0), np.float32(0.4073), np.float32(0.5168), np.float32(0.0), np.float32(0.0), np.float32(0.662), np.float32(0.6179), np.float32(0.7801), np.float32(0.7306)] 
+2026-10-16 22:53:48.081358: Epoch time: 228.12 s 
+2026-10-16 22:53:48.081555: Yayy! New best EMA pseudo Dice: 0.3725999891757965 
+2026-10-16 22:53:49.506368:  
+2026-10-16 22:53:49.507163: Epoch 62 
+2026-10-16 22:53:49.507602: Current learning rate: 0.00944 
+2026-10-16 22:57:46.908087: train_loss -0.5851 
+2026-10-16 22:57:46.908863: val_loss -0.357 
+2026-10-16 22:57:46.909348: Pseudo dice [np.float32(0.8759), np.float32(0.3694), np.float32(0.3607), np.float32(0.4595), np.float32(0.3231), np.float32(0.4068), np.float32(0.3945), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3839), np.float32(0.3798), np.float32(0.3766), np.float32(0.3724), np.float32(0.0), np.float32(0.0), np.float32(0.3778), np.float32(0.3166), np.float32(0.3292), np.float32(0.3411), np.float32(0.2443), np.float32(0.2487), np.float32(0.3207), np.float32(0.4674), np.float32(0.2291), np.float32(0.3103), np.float32(0.0), np.float32(0.0), np.float32(0.0237), np.float32(0.49), np.float32(0.0), np.float32(0.0), np.float32(0.3054), np.float32(0.2096), np.float32(0.3448), np.float32(0.3209)] 
+2026-10-16 22:57:46.910013: Epoch time: 237.41 s 
+2026-10-16 22:57:47.753922:  
+2026-10-16 22:57:47.754629: Epoch 63 
+2026-10-16 22:57:47.755018: Current learning rate: 0.00943 
+2026-10-16 23:01:23.360725: train_loss -0.5838 
+2026-10-16 23:01:23.361438: val_loss -0.486 
+2026-10-16 23:01:23.362019: Pseudo dice [np.float32(0.876), np.float32(0.619), np.float32(0.6434), np.float32(0.7768), np.float32(0.5818), np.float32(0.7525), np.float32(0.6231), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.649), np.float32(0.6177), np.float32(0.6412), np.float32(0.6053), np.float32(0.0), np.float32(0.0), np.float32(0.6325), np.float32(0.4914), np.float32(0.6178), np.float32(0.47), np.float32(0.4206), np.float32(0.5644), np.float32(0.4741), np.float32(0.6234), np.float32(0.4646), np.float32(0.4497), np.float32(0.0), np.float32(0.0), np.float32(0.4249), np.float32(0.5175), np.float32(0.0), np.float32(0.0), np.float32(0.5841), np.float32(0.5758), np.float32(0.7097), np.float32(0.6907)] 
+2026-10-16 23:01:23.362525: Epoch time: 215.61 s 
+2026-10-16 23:01:24.214844:  
+2026-10-16 23:01:24.215533: Epoch 64 
+2026-10-16 23:01:24.215870: Current learning rate: 0.00942 
+2026-10-16 23:05:22.294734: train_loss -0.5746 
+2026-10-16 23:05:22.295410: val_loss -0.3179 
+2026-10-16 23:05:22.296183: Pseudo dice [np.float32(0.8613), np.float32(0.2827), np.float32(0.3071), np.float32(0.3725), np.float32(0.2686), np.float32(0.3786), np.float32(0.2978), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3381), np.float32(0.341), np.float32(0.4299), np.float32(0.3736), np.float32(0.0), np.float32(0.0), np.float32(0.3718), np.float32(0.326), np.float32(0.2679), np.float32(0.2019), np.float32(0.0925), np.float32(0.2696), np.float32(0.453), np.float32(0.3853), np.float32(0.2825), np.float32(0.2672), np.float32(0.0), np.float32(0.0), np.float32(0.0606), np.float32(0.4001), np.float32(0.0), np.float32(0.0), np.float32(0.3038), np.float32(0.2819), np.float32(0.3299), np.float32(0.2986)] 
+2026-10-16 23:05:22.296730: Epoch time: 238.08 s 
+2026-10-16 23:05:23.119326:  
+2026-10-16 23:05:23.120073: Epoch 65 
+2026-10-16 23:05:23.120336: Current learning rate: 0.00941 
+2026-10-16 23:09:13.137861: train_loss -0.5298 
+2026-10-16 23:09:13.138765: val_loss -0.4527 
+2026-10-16 23:09:13.139308: Pseudo dice [np.float32(0.8704), np.float32(0.5835), np.float32(0.5756), np.float32(0.6937), np.float32(0.516), np.float32(0.6869), np.float32(0.5769), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6243), np.float32(0.5473), np.float32(0.5606), np.float32(0.5196), np.float32(0.0), np.float32(0.0), np.float32(0.5778), np.float32(0.4704), np.float32(0.5924), np.float32(0.4455), np.float32(0.5848), np.float32(0.4913), np.float32(0.3551), np.float32(0.5515), np.float32(0.4545), np.float32(0.3856), np.float32(0.0), np.float32(0.0), np.float32(0.3981), np.float32(0.5069), np.float32(0.0), np.float32(0.0), np.float32(0.5514), np.float32(0.5356), np.float32(0.6509), np.float32(0.6391)] 
+2026-10-16 23:09:13.139864: Epoch time: 230.02 s 
+2026-10-16 23:09:13.962355:  
+2026-10-16 23:09:13.963007: Epoch 66 
+2026-10-16 23:09:13.963441: Current learning rate: 0.0094 
+2026-10-16 23:13:13.512316: train_loss -0.5561 
+2026-10-16 23:13:13.512979: val_loss -0.337 
+2026-10-16 23:13:13.513787: Pseudo dice [np.float32(0.8788), np.float32(0.2845), np.float32(0.3317), np.float32(0.4115), np.float32(0.3392), np.float32(0.3674), np.float32(0.3575), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.372), np.float32(0.3673), np.float32(0.3792), np.float32(0.3752), np.float32(0.0), np.float32(0.0), np.float32(0.4227), np.float32(0.3397), np.float32(0.3126), np.float32(0.292), np.float32(0.1397), np.float32(0.2388), np.float32(0.4471), np.float32(0.4524), np.float32(0.3217), np.float32(0.2336), np.float32(0.0), np.float32(0.0), np.float32(0.0547), np.float32(0.4176), np.float32(0.0), np.float32(0.0), np.float32(0.2473), np.float32(0.1806), np.float32(0.3031), np.float32(0.3137)] 
+2026-10-16 23:13:13.514061: Epoch time: 239.55 s 
+2026-10-16 23:13:14.290139:  
+2026-10-16 23:13:14.290860: Epoch 67 
+2026-10-16 23:13:14.291216: Current learning rate: 0.00939 
+2026-10-16 23:17:02.680578: train_loss -0.5621 
+2026-10-16 23:17:02.681566: val_loss -0.4893 
+2026-10-16 23:17:02.682138: Pseudo dice [np.float32(0.8703), np.float32(0.6282), np.float32(0.6416), np.float32(0.7444), np.float32(0.6122), np.float32(0.7451), np.float32(0.6157), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6242), np.float32(0.6018), np.float32(0.6043), np.float32(0.604), np.float32(0.0), np.float32(0.0), np.float32(0.6255), np.float32(0.4784), np.float32(0.5892), np.float32(0.4717), np.float32(0.4575), np.float32(0.4657), np.float32(0.6547), np.float32(0.7009), np.float32(0.4397), np.float32(0.4631), np.float32(0.0), np.float32(0.0), np.float32(0.2321), np.float32(0.5502), np.float32(0.0), np.float32(0.0), np.float32(0.5746), np.float32(0.5606), np.float32(0.7091), np.float32(0.7096)] 
+2026-10-16 23:17:02.682444: Epoch time: 228.39 s 
+2026-10-16 23:17:03.860101:  
+2026-10-16 23:17:03.860657: Epoch 68 
+2026-10-16 23:17:03.861071: Current learning rate: 0.00939 
+2026-10-16 23:20:48.878408: train_loss -0.5847 
+2026-10-16 23:20:48.878941: val_loss -0.4317 
+2026-10-16 23:20:48.879780: Pseudo dice [np.float32(0.882), np.float32(0.5207), np.float32(0.5316), np.float32(0.6486), np.float32(0.4939), np.float32(0.6152), np.float32(0.5212), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5867), np.float32(0.5372), np.float32(0.5733), np.float32(0.5067), np.float32(0.0), np.float32(0.0), np.float32(0.5687), np.float32(0.4107), np.float32(0.5059), np.float32(0.4171), np.float32(0.4485), np.float32(0.4262), np.float32(0.4434), np.float32(0.4423), np.float32(0.4876), np.float32(0.3879), np.float32(0.0), np.float32(0.0), np.float32(0.2824), np.float32(0.4694), np.float32(0.0), np.float32(0.0), np.float32(0.5385), np.float32(0.519), np.float32(0.5529), np.float32(0.5457)] 
+2026-10-16 23:20:48.880230: Epoch time: 225.02 s 
+2026-10-16 23:20:49.730845:  
+2026-10-16 23:20:49.731744: Epoch 69 
+2026-10-16 23:20:49.732106: Current learning rate: 0.00938 
+2026-10-16 23:24:19.427784: train_loss -0.5798 
+2026-10-16 23:24:19.428438: val_loss -0.3658 
+2026-10-16 23:24:19.428885: Pseudo dice [np.float32(0.8856), np.float32(0.4253), np.float32(0.3691), np.float32(0.5014), np.float32(0.3806), np.float32(0.4591), np.float32(0.3865), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4247), np.float32(0.4319), np.float32(0.5174), np.float32(0.425), np.float32(0.0), np.float32(0.0), np.float32(0.4319), np.float32(0.316), np.float32(0.4016), np.float32(0.334), np.float32(0.2739), np.float32(0.3067), np.float32(0.3027), np.float32(0.4937), np.float32(0.2932), np.float32(0.3142), np.float32(0.0), np.float32(0.0), np.float32(0.0442), np.float32(0.4628), np.float32(0.0), np.float32(0.0), np.float32(0.3328), np.float32(0.3228), np.float32(0.4342), np.float32(0.4147)] 
+2026-10-16 23:24:19.429326: Epoch time: 209.7 s 
+2026-10-16 23:24:20.265458:  
+2026-10-16 23:24:20.266106: Epoch 70 
+2026-10-16 23:24:20.266727: Current learning rate: 0.00937 
+2026-10-16 23:28:04.532362: train_loss -0.5842 
+2026-10-16 23:28:04.532903: val_loss -0.354 
+2026-10-16 23:28:04.533539: Pseudo dice [np.float32(0.8673), np.float32(0.3554), np.float32(0.4002), np.float32(0.4468), np.float32(0.3413), np.float32(0.3813), np.float32(0.3398), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3645), np.float32(0.374), np.float32(0.4251), np.float32(0.3632), np.float32(0.0), np.float32(0.0), np.float32(0.4283), np.float32(0.2868), np.float32(0.3777), np.float32(0.3382), np.float32(0.0615), np.float32(0.3171), np.float32(0.3506), np.float32(0.5882), np.float32(0.3089), np.float32(0.303), np.float32(0.0), np.float32(0.0), np.float32(0.0438), np.float32(0.4158), np.float32(0.0), np.float32(0.0), np.float32(0.3367), np.float32(0.209), np.float32(0.3769), np.float32(0.4084)] 
+2026-10-16 23:28:04.533784: Epoch time: 224.27 s 
+2026-10-16 23:28:05.370405:  
+2026-10-16 23:28:05.371113: Epoch 71 
+2026-10-16 23:28:05.371643: Current learning rate: 0.00936 
+2026-10-16 23:32:01.575267: train_loss -0.5844 
+2026-10-16 23:32:01.576082: val_loss -0.3922 
+2026-10-16 23:32:01.576798: Pseudo dice [np.float32(0.837), np.float32(0.4527), np.float32(0.4567), np.float32(0.6021), np.float32(0.4697), np.float32(0.5693), np.float32(0.4252), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5204), np.float32(0.4882), np.float32(0.5613), np.float32(0.4651), np.float32(0.0), np.float32(0.0), np.float32(0.5329), np.float32(0.3938), np.float32(0.4469), np.float32(0.2772), np.float32(0.4051), np.float32(0.4451), np.float32(0.3685), np.float32(0.3347), np.float32(0.4375), np.float32(0.3827), np.float32(0.0), np.float32(0.0), np.float32(0.3968), np.float32(0.4072), np.float32(0.0), np.float32(0.0), np.float32(0.477), np.float32(0.4098), np.float32(0.4786), np.float32(0.4678)] 
+2026-10-16 23:32:01.577042: Epoch time: 236.21 s 
+2026-10-16 23:32:02.509045:  
+2026-10-16 23:32:02.509807: Epoch 72 
+2026-10-16 23:32:02.510159: Current learning rate: 0.00935 
+2026-10-16 23:35:34.935628: train_loss -0.5811 
+2026-10-16 23:35:34.936489: val_loss -0.4125 
+2026-10-16 23:35:34.937141: Pseudo dice [np.float32(0.8205), np.float32(0.4797), np.float32(0.4648), np.float32(0.591), np.float32(0.4773), np.float32(0.5719), np.float32(0.5003), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4941), np.float32(0.4471), np.float32(0.4848), np.float32(0.4206), np.float32(0.0), np.float32(0.0), np.float32(0.5248), np.float32(0.4023), np.float32(0.4479), np.float32(0.3208), np.float32(0.3181), np.float32(0.3316), np.float32(0.5536), np.float32(0.5781), np.float32(0.3086), np.float32(0.3964), np.float32(0.0), np.float32(0.0), np.float32(0.0687), np.float32(0.4956), np.float32(0.0), np.float32(0.0), np.float32(0.4176), np.float32(0.4316), np.float32(0.5411), np.float32(0.4871)] 
+2026-10-16 23:35:34.937534: Epoch time: 212.43 s 
+2026-10-16 23:35:35.737532:  
+2026-10-16 23:35:35.738336: Epoch 73 
+2026-10-16 23:35:35.738708: Current learning rate: 0.00934 
+2026-10-16 23:39:08.151512: train_loss -0.5707 
+2026-10-16 23:39:08.152438: val_loss -0.338 
+2026-10-16 23:39:08.153006: Pseudo dice [np.float32(0.8303), np.float32(0.3984), np.float32(0.3821), np.float32(0.3551), np.float32(0.312), np.float32(0.3184), np.float32(0.28), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3602), np.float32(0.3821), np.float32(0.3772), np.float32(0.339), np.float32(0.0), np.float32(0.0), np.float32(0.3762), np.float32(0.3447), np.float32(0.2576), np.float32(0.304), np.float32(0.3036), np.float32(0.3078), np.float32(0.4341), np.float32(0.4844), np.float32(0.3244), np.float32(0.405), np.float32(0.0), np.float32(0.0), np.float32(0.032), np.float32(0.4739), np.float32(0.0), np.float32(0.0), np.float32(0.2621), np.float32(0.1587), np.float32(0.3058), np.float32(0.3217)] 
+2026-10-16 23:39:08.153435: Epoch time: 212.42 s 
+2026-10-16 23:39:09.008111:  
+2026-10-16 23:39:09.008728: Epoch 74 
+2026-10-16 23:39:09.009103: Current learning rate: 0.00933 
+2026-10-16 23:42:59.401162: train_loss -0.5752 
+2026-10-16 23:42:59.401863: val_loss -0.3002 
+2026-10-16 23:42:59.402445: Pseudo dice [np.float32(0.8577), np.float32(0.298), np.float32(0.2641), np.float32(0.249), np.float32(0.226), np.float32(0.3002), np.float32(0.324), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3079), np.float32(0.3427), np.float32(0.3221), np.float32(0.2453), np.float32(0.0), np.float32(0.0), np.float32(0.3304), np.float32(0.2345), np.float32(0.2362), np.float32(0.2031), np.float32(0.2234), np.float32(0.2177), np.float32(0.2289), np.float32(0.097), np.float32(0.2109), np.float32(0.2893), np.float32(0.0), np.float32(0.0), np.float32(0.0382), np.float32(0.1454), np.float32(0.0), np.float32(0.0), np.float32(0.1769), np.float32(0.1637), np.float32(0.2665), np.float32(0.2362)] 
+2026-10-16 23:42:59.402903: Epoch time: 230.4 s 
+2026-10-16 23:43:00.263034:  
+2026-10-16 23:43:00.263604: Epoch 75 
+2026-10-16 23:43:00.264045: Current learning rate: 0.00932 
+2026-10-16 23:46:39.743323: train_loss -0.5806 
+2026-10-16 23:46:39.744488: val_loss -0.5058 
+2026-10-16 23:46:39.745044: Pseudo dice [np.float32(0.8868), np.float32(0.7166), np.float32(0.6624), np.float32(0.7495), np.float32(0.6442), np.float32(0.7533), np.float32(0.6947), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6114), np.float32(0.6167), np.float32(0.5185), np.float32(0.4832), np.float32(0.0), np.float32(0.0), np.float32(0.6614), np.float32(0.4774), np.float32(0.6593), np.float32(0.5241), np.float32(0.5434), np.float32(0.528), np.float32(0.7242), np.float32(0.77), np.float32(0.5094), np.float32(0.5039), np.float32(0.0), np.float32(0.0), np.float32(0.1418), np.float32(0.5557), np.float32(0.0), np.float32(0.0), np.float32(0.5949), np.float32(0.5251), np.float32(0.7348), np.float32(0.7173)] 
+2026-10-16 23:46:39.745707: Epoch time: 219.48 s 
+2026-10-16 23:46:40.677961:  
+2026-10-16 23:46:40.678730: Epoch 76 
+2026-10-16 23:46:40.678953: Current learning rate: 0.00931 
+2026-10-16 23:50:35.587255: train_loss -0.6006 
+2026-10-16 23:50:35.588302: val_loss -0.4319 
+2026-10-16 23:50:35.588905: Pseudo dice [np.float32(0.866), np.float32(0.5548), np.float32(0.5428), np.float32(0.6417), np.float32(0.4817), np.float32(0.6377), np.float32(0.5154), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5786), np.float32(0.5191), np.float32(0.5977), np.float32(0.482), np.float32(0.0), np.float32(0.0), np.float32(0.5103), np.float32(0.4278), np.float32(0.5369), np.float32(0.411), np.float32(0.4811), np.float32(0.4275), np.float32(0.5069), np.float32(0.5163), np.float32(0.3284), np.float32(0.2873), np.float32(0.0), np.float32(0.0), np.float32(0.2981), np.float32(0.4938), np.float32(0.0), np.float32(0.0), np.float32(0.4458), np.float32(0.4616), np.float32(0.5782), np.float32(0.5776)] 
+2026-10-16 23:50:35.589468: Epoch time: 234.91 s 
+2026-10-16 23:50:36.463590:  
+2026-10-16 23:50:36.464285: Epoch 77 
+2026-10-16 23:50:36.464673: Current learning rate: 0.0093 
+2026-10-16 23:54:30.952569: train_loss -0.6098 
+2026-10-16 23:54:30.953120: val_loss -0.3841 
+2026-10-16 23:54:30.953755: Pseudo dice [np.float32(0.8615), np.float32(0.4528), np.float32(0.4191), np.float32(0.4572), np.float32(0.368), np.float32(0.4906), np.float32(0.45), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3982), np.float32(0.4367), np.float32(0.4944), np.float32(0.3739), np.float32(0.0), np.float32(0.0), np.float32(0.4476), np.float32(0.3723), np.float32(0.4587), np.float32(0.4254), np.float32(0.1854), np.float32(0.2065), np.float32(0.4046), np.float32(0.5155), np.float32(0.2635), np.float32(0.3875), np.float32(0.0), np.float32(0.0), np.float32(0.0391), np.float32(0.424), np.float32(0.0), np.float32(0.0), np.float32(0.4089), np.float32(0.29), np.float32(0.3691), np.float32(0.4405)] 
+2026-10-16 23:54:30.954865: Epoch time: 234.49 s 
+2026-10-16 23:54:31.845495:  
+2026-10-16 23:54:31.846082: Epoch 78 
+2026-10-16 23:54:31.846632: Current learning rate: 0.0093 
+2026-10-16 23:58:25.051986: train_loss -0.6003 
+2026-10-16 23:58:25.053119: val_loss -0.2425 
+2026-10-16 23:58:25.054149: Pseudo dice [np.float32(0.7808), np.float32(0.1824), np.float32(0.16), np.float32(0.2082), np.float32(0.1644), np.float32(0.2255), np.float32(0.1938), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2248), np.float32(0.2574), np.float32(0.2618), np.float32(0.2047), np.float32(0.0), np.float32(0.0), np.float32(0.2511), np.float32(0.189), np.float32(0.1772), np.float32(0.1595), np.float32(0.1303), np.float32(0.1543), np.float32(0.1577), np.float32(0.3082), np.float32(0.1529), np.float32(0.2592), np.float32(0.0), np.float32(0.0), np.float32(0.0446), np.float32(0.21), np.float32(0.0), np.float32(0.0), np.float32(0.1161), np.float32(0.1444), np.float32(0.1658), np.float32(0.2077)] 
+2026-10-16 23:58:25.054811: Epoch time: 233.21 s 
+2026-10-16 23:58:25.889348:  
+2026-10-16 23:58:25.890069: Epoch 79 
+2026-10-16 23:58:25.890342: Current learning rate: 0.00929 
+2026-10-17 00:02:09.595750: train_loss -0.5934 
+2026-10-17 00:02:09.596674: val_loss -0.2272 
+2026-10-17 00:02:09.597562: Pseudo dice [np.float32(0.8799), np.float32(0.1212), np.float32(0.1649), np.float32(0.0889), np.float32(0.0885), np.float32(0.1743), np.float32(0.2295), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1303), np.float32(0.2136), np.float32(0.2332), np.float32(0.1317), np.float32(0.0), np.float32(0.0), np.float32(0.1856), np.float32(0.143), np.float32(0.2031), np.float32(0.1294), np.float32(0.1094), np.float32(0.1265), np.float32(0.2445), np.float32(0.2346), np.float32(0.1955), np.float32(0.2018), np.float32(0.0), np.float32(0.0), np.float32(0.0876), np.float32(0.2362), np.float32(0.0), np.float32(0.0), np.float32(0.1038), np.float32(0.0967), np.float32(0.1332), np.float32(0.1961)] 
+2026-10-17 00:02:09.598223: Epoch time: 223.71 s 
+2026-10-17 00:02:10.462696:  
+2026-10-17 00:02:10.463407: Epoch 80 
+2026-10-17 00:02:10.463817: Current learning rate: 0.00928 
+2026-10-17 00:05:44.942916: train_loss -0.5896 
+2026-10-17 00:05:44.943527: val_loss -0.375 
+2026-10-17 00:05:44.944123: Pseudo dice [np.float32(0.8763), np.float32(0.4055), np.float32(0.4432), np.float32(0.5281), np.float32(0.4153), np.float32(0.4775), np.float32(0.4234), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.47), np.float32(0.4327), np.float32(0.4833), np.float32(0.4256), np.float32(0.0), np.float32(0.0), np.float32(0.5199), np.float32(0.3824), np.float32(0.5038), np.float32(0.3778), np.float32(0.4902), np.float32(0.3773), np.float32(0.2901), np.float32(0.5035), np.float32(0.3876), np.float32(0.2959), np.float32(0.0), np.float32(0.0), np.float32(0.3066), np.float32(0.4629), np.float32(0.0), np.float32(0.0), np.float32(0.4602), np.float32(0.3704), np.float32(0.4577), np.float32(0.4685)] 
+2026-10-17 00:05:44.944767: Epoch time: 214.48 s 
+2026-10-17 00:05:45.817262:  
+2026-10-17 00:05:45.817734: Epoch 81 
+2026-10-17 00:05:45.818577: Current learning rate: 0.00927 
+2026-10-17 00:09:39.203762: train_loss -0.5864 
+2026-10-17 00:09:39.204423: val_loss -0.373 
+2026-10-17 00:09:39.205012: Pseudo dice [np.float32(0.8242), np.float32(0.4333), np.float32(0.4126), np.float32(0.4198), np.float32(0.3816), np.float32(0.4926), np.float32(0.4762), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3681), np.float32(0.4506), np.float32(0.4073), np.float32(0.1089), np.float32(0.0), np.float32(0.0), np.float32(0.464), np.float32(0.3239), np.float32(0.3434), np.float32(0.2141), np.float32(0.3362), np.float32(0.3272), np.float32(0.5794), np.float32(0.6249), np.float32(0.3256), np.float32(0.3389), np.float32(0.0), np.float32(0.0), np.float32(0.2357), np.float32(0.4258), np.float32(0.0), np.float32(0.0), np.float32(0.2996), np.float32(0.3868), np.float32(0.5953), np.float32(0.5884)] 
+2026-10-17 00:09:39.205740: Epoch time: 233.39 s 
+2026-10-17 00:09:40.009467:  
+2026-10-17 00:09:40.010239: Epoch 82 
+2026-10-17 00:09:40.010685: Current learning rate: 0.00926 
+2026-10-17 00:13:18.428485: train_loss -0.5886 
+2026-10-17 00:13:18.429362: val_loss -0.3313 
+2026-10-17 00:13:18.429930: Pseudo dice [np.float32(0.8895), np.float32(0.3085), np.float32(0.2812), np.float32(0.3576), np.float32(0.3038), np.float32(0.356), np.float32(0.3612), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2918), np.float32(0.3352), np.float32(0.3356), np.float32(0.2992), np.float32(0.0), np.float32(0.0), np.float32(0.3831), np.float32(0.2978), np.float32(0.2815), np.float32(0.2765), np.float32(0.1635), np.float32(0.2206), np.float32(0.5357), np.float32(0.4181), np.float32(0.2843), np.float32(0.371), np.float32(0.0), np.float32(0.0), np.float32(0.0465), np.float32(0.3512), np.float32(0.0), np.float32(0.0), np.float32(0.2406), np.float32(0.2517), np.float32(0.3402), np.float32(0.3455)] 
+2026-10-17 00:13:18.430207: Epoch time: 218.42 s 
+2026-10-17 00:13:19.220421:  
+2026-10-17 00:13:19.221084: Epoch 83 
+2026-10-17 00:13:19.221323: Current learning rate: 0.00925 
+2026-10-17 00:16:51.181765: train_loss -0.5851 
+2026-10-17 00:16:51.182782: val_loss -0.4225 
+2026-10-17 00:16:51.183344: Pseudo dice [np.float32(0.8646), np.float32(0.5083), np.float32(0.4584), np.float32(0.6075), np.float32(0.4601), np.float32(0.5866), np.float32(0.5239), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5387), np.float32(0.5446), np.float32(0.505), np.float32(0.4243), np.float32(0.0), np.float32(0.0), np.float32(0.4841), np.float32(0.411), np.float32(0.4397), np.float32(0.3251), np.float32(0.2239), np.float32(0.2442), np.float32(0.4565), np.float32(0.5234), np.float32(0.3389), np.float32(0.3833), np.float32(0.0), np.float32(0.0), np.float32(0.1272), np.float32(0.3706), np.float32(0.0), np.float32(0.0), np.float32(0.4689), np.float32(0.426), np.float32(0.5034), np.float32(0.4888)] 
+2026-10-17 00:16:51.183829: Epoch time: 211.96 s 
+2026-10-17 00:16:51.956289:  
+2026-10-17 00:16:51.956866: Epoch 84 
+2026-10-17 00:16:51.957246: Current learning rate: 0.00924 
+2026-10-17 00:20:38.781753: train_loss -0.5932 
+2026-10-17 00:20:38.782462: val_loss -0.3749 
+2026-10-17 00:20:38.782961: Pseudo dice [np.float32(0.8596), np.float32(0.3624), np.float32(0.3911), np.float32(0.4595), np.float32(0.3514), np.float32(0.4167), np.float32(0.383), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3784), np.float32(0.3805), np.float32(0.3755), np.float32(0.3821), np.float32(0.0), np.float32(0.0), np.float32(0.4192), np.float32(0.3144), np.float32(0.3252), np.float32(0.3139), np.float32(0.2327), np.float32(0.2778), np.float32(0.3239), np.float32(0.5951), np.float32(0.3103), np.float32(0.3234), np.float32(0.0), np.float32(0.0), np.float32(0.0732), np.float32(0.5195), np.float32(0.0), np.float32(0.0), np.float32(0.2823), np.float32(0.2365), np.float32(0.3728), np.float32(0.3925)] 
+2026-10-17 00:20:38.783534: Epoch time: 226.83 s 
+2026-10-17 00:20:39.628932:  
+2026-10-17 00:20:39.629772: Epoch 85 
+2026-10-17 00:20:39.630203: Current learning rate: 0.00923 
+2026-10-17 00:24:12.194635: train_loss -0.6151 
+2026-10-17 00:24:12.195662: val_loss -0.2942 
+2026-10-17 00:24:12.196144: Pseudo dice [np.float32(0.8493), np.float32(0.2579), np.float32(0.2552), np.float32(0.2866), np.float32(0.21), np.float32(0.3005), np.float32(0.2269), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2888), np.float32(0.2956), np.float32(0.2772), np.float32(0.2406), np.float32(0.0), np.float32(0.0), np.float32(0.3525), np.float32(0.2964), np.float32(0.2041), np.float32(0.1934), np.float32(0.2312), np.float32(0.1585), np.float32(0.3998), np.float32(0.3097), np.float32(0.2357), np.float32(0.349), np.float32(0.0), np.float32(0.0), np.float32(0.0605), np.float32(0.2888), np.float32(0.0), np.float32(0.0), np.float32(0.1395), np.float32(0.1317), np.float32(0.2451), np.float32(0.2605)] 
+2026-10-17 00:24:12.196591: Epoch time: 212.57 s 
+2026-10-17 00:24:13.315171:  
+2026-10-17 00:24:13.316077: Epoch 86 
+2026-10-17 00:24:13.316589: Current learning rate: 0.00922 
+2026-10-17 00:28:17.501601: train_loss -0.598 
+2026-10-17 00:28:17.502504: val_loss -0.3289 
+2026-10-17 00:28:17.503083: Pseudo dice [np.float32(0.8179), np.float32(0.2944), np.float32(0.329), np.float32(0.3945), np.float32(0.2805), np.float32(0.3473), np.float32(0.3153), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.369), np.float32(0.3713), np.float32(0.3454), np.float32(0.3339), np.float32(0.0), np.float32(0.0), np.float32(0.3507), np.float32(0.2617), np.float32(0.316), np.float32(0.2553), np.float32(0.2417), np.float32(0.2728), np.float32(0.306), np.float32(0.5), np.float32(0.3289), np.float32(0.2881), np.float32(0.0), np.float32(0.0), np.float32(0.1061), np.float32(0.4525), np.float32(0.0), np.float32(0.0), np.float32(0.229), np.float32(0.1939), np.float32(0.2945), np.float32(0.335)] 
+2026-10-17 00:28:17.503555: Epoch time: 244.19 s 
+2026-10-17 00:28:18.304921:  
+2026-10-17 00:28:18.305557: Epoch 87 
+2026-10-17 00:28:18.306009: Current learning rate: 0.00921 
+2026-10-17 00:31:53.761182: train_loss -0.6118 
+2026-10-17 00:31:53.762175: val_loss -0.3888 
+2026-10-17 00:31:53.762929: Pseudo dice [np.float32(0.8761), np.float32(0.439), np.float32(0.452), np.float32(0.5639), np.float32(0.4033), np.float32(0.5495), np.float32(0.4355), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5082), np.float32(0.469), np.float32(0.4982), np.float32(0.3896), np.float32(0.0), np.float32(0.0), np.float32(0.4957), np.float32(0.4125), np.float32(0.4605), np.float32(0.3621), np.float32(0.3741), np.float32(0.3666), np.float32(0.3524), np.float32(0.4691), np.float32(0.4228), np.float32(0.3553), np.float32(0.0), np.float32(0.0), np.float32(0.2611), np.float32(0.4746), np.float32(0.0), np.float32(0.0), np.float32(0.4946), np.float32(0.4621), np.float32(0.4936), np.float32(0.4803)] 
+2026-10-17 00:31:53.763373: Epoch time: 215.46 s 
+2026-10-17 00:31:54.529537:  
+2026-10-17 00:31:54.530264: Epoch 88 
+2026-10-17 00:31:54.530677: Current learning rate: 0.0092 
+2026-10-17 00:35:41.246503: train_loss -0.6214 
+2026-10-17 00:35:41.247594: val_loss -0.3138 
+2026-10-17 00:35:41.248397: Pseudo dice [np.float32(0.8653), np.float32(0.2858), np.float32(0.2916), np.float32(0.3243), np.float32(0.2463), np.float32(0.3074), np.float32(0.2767), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2844), np.float32(0.2713), np.float32(0.3504), np.float32(0.3347), np.float32(0.0), np.float32(0.0), np.float32(0.3538), np.float32(0.3335), np.float32(0.2382), np.float32(0.2169), np.float32(0.2895), np.float32(0.258), np.float32(0.1908), np.float32(0.2805), np.float32(0.2096), np.float32(0.2698), np.float32(0.0), np.float32(0.0), np.float32(0.0667), np.float32(0.3253), np.float32(0.0), np.float32(0.0), np.float32(0.2078), np.float32(0.2564), np.float32(0.2555), np.float32(0.2527)] 
+2026-10-17 00:35:41.248757: Epoch time: 226.72 s 
+2026-10-17 00:35:42.119188:  
+2026-10-17 00:35:42.119929: Epoch 89 
+2026-10-17 00:35:42.120428: Current learning rate: 0.0092 
+2026-10-17 00:39:23.727119: train_loss -0.6258 
+2026-10-17 00:39:23.727800: val_loss -0.5162 
+2026-10-17 00:39:23.728589: Pseudo dice [np.float32(0.8578), np.float32(0.6357), np.float32(0.6104), np.float32(0.7258), np.float32(0.6266), np.float32(0.7172), np.float32(0.6631), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6095), np.float32(0.5727), np.float32(0.5738), np.float32(0.5281), np.float32(0.0), np.float32(0.0), np.float32(0.6735), np.float32(0.5507), np.float32(0.6373), np.float32(0.5355), np.float32(0.5979), np.float32(0.5303), np.float32(0.6472), np.float32(0.6216), np.float32(0.4558), np.float32(0.4827), np.float32(0.0), np.float32(0.0), np.float32(0.3884), np.float32(0.5114), np.float32(0.0), np.float32(0.0), np.float32(0.6361), np.float32(0.6024), np.float32(0.6997), np.float32(0.6796)] 
+2026-10-17 00:39:23.729427: Epoch time: 221.61 s 
+2026-10-17 00:39:24.500087:  
+2026-10-17 00:39:24.500825: Epoch 90 
+2026-10-17 00:39:24.501260: Current learning rate: 0.00919 
+2026-10-17 00:43:16.269379: train_loss -0.6272 
+2026-10-17 00:43:16.270438: val_loss -0.3686 
+2026-10-17 00:43:16.270897: Pseudo dice [np.float32(0.7912), np.float32(0.3586), np.float32(0.3829), np.float32(0.4365), np.float32(0.4303), np.float32(0.4455), np.float32(0.3295), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3936), np.float32(0.4327), np.float32(0.2496), np.float32(0.3533), np.float32(0.0), np.float32(0.0), np.float32(0.4955), np.float32(0.3993), np.float32(0.4758), np.float32(0.4132), np.float32(0.3771), np.float32(0.3953), np.float32(0.2108), np.float32(0.3396), np.float32(0.342), np.float32(0.3205), np.float32(0.0), np.float32(0.0), np.float32(0.3416), np.float32(0.3348), np.float32(0.0), np.float32(0.0), np.float32(0.2532), np.float32(0.3638), np.float32(0.4485), np.float32(0.4516)] 
+2026-10-17 00:43:16.271390: Epoch time: 231.77 s 
+2026-10-17 00:43:17.092322:  
+2026-10-17 00:43:17.093181: Epoch 91 
+2026-10-17 00:43:17.093475: Current learning rate: 0.00918 
+2026-10-17 00:47:08.653330: train_loss -0.6226 
+2026-10-17 00:47:08.653971: val_loss -0.4898 
+2026-10-17 00:47:08.654551: Pseudo dice [np.float32(0.8617), np.float32(0.5987), np.float32(0.5927), np.float32(0.7027), np.float32(0.5575), np.float32(0.6732), np.float32(0.5737), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6126), np.float32(0.5568), np.float32(0.5547), np.float32(0.5106), np.float32(0.0), np.float32(0.0), np.float32(0.628), np.float32(0.5181), np.float32(0.6018), np.float32(0.4885), np.float32(0.545), np.float32(0.5029), np.float32(0.5428), np.float32(0.5074), np.float32(0.4835), np.float32(0.4274), np.float32(0.0), np.float32(0.0), np.float32(0.4173), np.float32(0.4157), np.float32(0.0), np.float32(0.0), np.float32(0.5432), np.float32(0.4946), np.float32(0.6307), np.float32(0.5957)] 
+2026-10-17 00:47:08.655054: Epoch time: 231.56 s 
+2026-10-17 00:47:09.485698:  
+2026-10-17 00:47:09.486763: Epoch 92 
+2026-10-17 00:47:09.487022: Current learning rate: 0.00917 
+2026-10-17 00:50:55.815777: train_loss -0.6317 
+2026-10-17 00:50:55.816507: val_loss -0.3695 
+2026-10-17 00:50:55.817227: Pseudo dice [np.float32(0.87), np.float32(0.3669), np.float32(0.3736), np.float32(0.4127), np.float32(0.2954), np.float32(0.3995), np.float32(0.3421), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4017), np.float32(0.3686), np.float32(0.4322), np.float32(0.3051), np.float32(0.0), np.float32(0.0), np.float32(0.3638), np.float32(0.3067), np.float32(0.3183), np.float32(0.2733), np.float32(0.2112), np.float32(0.2862), np.float32(0.3224), np.float32(0.5163), np.float32(0.334), np.float32(0.2839), np.float32(0.0), np.float32(0.0), np.float32(0.0357), np.float32(0.4089), np.float32(0.0), np.float32(0.0), np.float32(0.268), np.float32(0.25), np.float32(0.3186), np.float32(0.3251)] 
+2026-10-17 00:50:55.817513: Epoch time: 226.33 s 
+2026-10-17 00:50:56.557994:  
+2026-10-17 00:50:56.558627: Epoch 93 
+2026-10-17 00:50:56.558966: Current learning rate: 0.00916 
+2026-10-17 00:54:37.880653: train_loss -0.6401 
+2026-10-17 00:54:37.881374: val_loss -0.4513 
+2026-10-17 00:54:37.881824: Pseudo dice [np.float32(0.7934), np.float32(0.5092), np.float32(0.5379), np.float32(0.6234), np.float32(0.488), np.float32(0.5931), np.float32(0.5085), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5455), np.float32(0.5239), np.float32(0.5573), np.float32(0.5031), np.float32(0.0), np.float32(0.0), np.float32(0.5209), np.float32(0.4187), np.float32(0.4993), np.float32(0.4017), np.float32(0.3919), np.float32(0.3966), np.float32(0.4283), np.float32(0.4662), np.float32(0.3965), np.float32(0.373), np.float32(0.0), np.float32(0.0), np.float32(0.3952), np.float32(0.4559), np.float32(0.0), np.float32(0.0), np.float32(0.5411), np.float32(0.481), np.float32(0.5665), np.float32(0.5524)] 
+2026-10-17 00:54:37.882345: Epoch time: 221.32 s 
+2026-10-17 00:54:38.668779:  
+2026-10-17 00:54:38.669521: Epoch 94 
+2026-10-17 00:54:38.669986: Current learning rate: 0.00915 
+2026-10-17 00:58:51.805177: train_loss -0.6127 
+2026-10-17 00:58:51.806090: val_loss -0.5385 
+2026-10-17 00:58:51.806735: Pseudo dice [np.float32(0.856), np.float32(0.7082), np.float32(0.677), np.float32(0.8029), np.float32(0.658), np.float32(0.7799), np.float32(0.7027), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7064), np.float32(0.6838), np.float32(0.6216), np.float32(0.6025), np.float32(0.0), np.float32(0.0), np.float32(0.6889), np.float32(0.5658), np.float32(0.676), np.float32(0.549), np.float32(0.4976), np.float32(0.486), np.float32(0.6799), np.float32(0.6711), np.float32(0.5338), np.float32(0.4124), np.float32(0.0), np.float32(0.0), np.float32(0.3053), np.float32(0.5532), np.float32(0.0), np.float32(0.0), np.float32(0.6726), np.float32(0.6112), np.float32(0.7529), np.float32(0.7364)] 
+2026-10-17 00:58:51.807259: Epoch time: 253.14 s 
+2026-10-17 00:58:52.637287:  
+2026-10-17 00:58:52.638146: Epoch 95 
+2026-10-17 00:58:52.638641: Current learning rate: 0.00914 
+2026-10-17 01:02:24.488124: train_loss -0.6084 
+2026-10-17 01:02:24.488713: val_loss -0.5608 
+2026-10-17 01:02:24.489275: Pseudo dice [np.float32(0.8904), np.float32(0.7467), np.float32(0.6716), np.float32(0.8433), np.float32(0.6784), np.float32(0.8221), np.float32(0.72), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7002), np.float32(0.7206), np.float32(0.6764), np.float32(0.6883), np.float32(0.0), np.float32(0.0), np.float32(0.7088), np.float32(0.5441), np.float32(0.7013), np.float32(0.5801), np.float32(0.5588), np.float32(0.5263), np.float32(0.7528), np.float32(0.7524), np.float32(0.4561), np.float32(0.5095), np.float32(0.0), np.float32(0.0), np.float32(0.3678), np.float32(0.5692), np.float32(0.0), np.float32(0.0), np.float32(0.6949), np.float32(0.6403), np.float32(0.8267), np.float32(0.8374)] 
+2026-10-17 01:02:24.489995: Epoch time: 211.85 s 
+2026-10-17 01:02:25.290794:  
+2026-10-17 01:02:25.291337: Epoch 96 
+2026-10-17 01:02:25.291702: Current learning rate: 0.00913 
+2026-10-17 01:06:33.027885: train_loss -0.57 
+2026-10-17 01:06:33.028824: val_loss -0.4182 
+2026-10-17 01:06:33.029419: Pseudo dice [np.float32(0.8539), np.float32(0.4404), np.float32(0.4715), np.float32(0.552), np.float32(0.4605), np.float32(0.5317), np.float32(0.4264), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5071), np.float32(0.4819), np.float32(0.4836), np.float32(0.4378), np.float32(0.0), np.float32(0.0), np.float32(0.5378), np.float32(0.3846), np.float32(0.4679), np.float32(0.381), np.float32(0.4951), np.float32(0.4039), np.float32(0.4443), np.float32(0.6067), np.float32(0.4183), np.float32(0.3701), np.float32(0.0), np.float32(0.0), np.float32(0.3989), np.float32(0.4682), np.float32(0.0), np.float32(0.0), np.float32(0.4575), np.float32(0.4245), np.float32(0.5464), np.float32(0.5139)] 
+2026-10-17 01:06:33.029859: Epoch time: 247.74 s 
+2026-10-17 01:06:33.808148:  
+2026-10-17 01:06:33.808845: Epoch 97 
+2026-10-17 01:06:33.809161: Current learning rate: 0.00912 
+2026-10-17 01:10:10.117279: train_loss -0.5875 
+2026-10-17 01:10:10.117977: val_loss -0.3599 
+2026-10-17 01:10:10.118515: Pseudo dice [np.float32(0.8466), np.float32(0.3311), np.float32(0.3838), np.float32(0.5183), np.float32(0.365), np.float32(0.4938), np.float32(0.3656), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4309), np.float32(0.4048), np.float32(0.4961), np.float32(0.4319), np.float32(0.0), np.float32(0.0), np.float32(0.3446), np.float32(0.2501), np.float32(0.4058), np.float32(0.2996), np.float32(0.3186), np.float32(0.3389), np.float32(0.1891), np.float32(0.3264), np.float32(0.2126), np.float32(0.1149), np.float32(0.0), np.float32(0.0), np.float32(0.3729), np.float32(0.2725), np.float32(0.0), np.float32(0.0), np.float32(0.4321), np.float32(0.422), np.float32(0.3851), np.float32(0.3727)] 
+2026-10-17 01:10:10.119038: Epoch time: 216.31 s 
+2026-10-17 01:10:11.015904:  
+2026-10-17 01:10:11.016633: Epoch 98 
+2026-10-17 01:10:11.016973: Current learning rate: 0.00911 
+2026-10-17 01:13:47.490683: train_loss -0.6186 
+2026-10-17 01:13:47.491445: val_loss -0.3515 
+2026-10-17 01:13:47.492198: Pseudo dice [np.float32(0.8249), np.float32(0.3428), np.float32(0.3697), np.float32(0.4444), np.float32(0.3195), np.float32(0.3931), np.float32(0.3486), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3272), np.float32(0.2927), np.float32(0.3929), np.float32(0.3255), np.float32(0.0), np.float32(0.0), np.float32(0.3493), np.float32(0.2187), np.float32(0.3186), np.float32(0.2399), np.float32(0.3334), np.float32(0.3028), np.float32(0.2686), np.float32(0.3003), np.float32(0.2848), np.float32(0.076), np.float32(0.0), np.float32(0.0), np.float32(0.4046), np.float32(0.2771), np.float32(0.0), np.float32(0.0), np.float32(0.3888), np.float32(0.3786), np.float32(0.3793), np.float32(0.3401)] 
+2026-10-17 01:13:47.492901: Epoch time: 216.48 s 
+2026-10-17 01:13:48.314421:  
+2026-10-17 01:13:48.314996: Epoch 99 
+2026-10-17 01:13:48.315641: Current learning rate: 0.0091 
+2026-10-17 01:17:30.497427: train_loss -0.5788 
+2026-10-17 01:17:30.498420: val_loss -0.4875 
+2026-10-17 01:17:30.499119: Pseudo dice [np.float32(0.8641), np.float32(0.5621), np.float32(0.5579), np.float32(0.7724), np.float32(0.5864), np.float32(0.744), np.float32(0.66), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5899), np.float32(0.5862), np.float32(0.5201), np.float32(0.4948), np.float32(0.0), np.float32(0.0), np.float32(0.6234), np.float32(0.4652), np.float32(0.5821), np.float32(0.4631), np.float32(0.5148), np.float32(0.4748), np.float32(0.5803), np.float32(0.5623), np.float32(0.5057), np.float32(0.3627), np.float32(0.0), np.float32(0.0), np.float32(0.3744), np.float32(0.4238), np.float32(0.0), np.float32(0.0), np.float32(0.6204), np.float32(0.5548), np.float32(0.6792), np.float32(0.6886)] 
+2026-10-17 01:17:30.499713: Epoch time: 222.19 s 
+2026-10-17 01:17:32.597733:  
+2026-10-17 01:17:32.598497: Epoch 100 
+2026-10-17 01:17:32.598828: Current learning rate: 0.0091 
+2026-10-17 01:21:13.185643: train_loss -0.6018 
+2026-10-17 01:21:13.186600: val_loss -0.5874 
+2026-10-17 01:21:13.187409: Pseudo dice [np.float32(0.8738), np.float32(0.7474), np.float32(0.6933), np.float32(0.8667), np.float32(0.7315), np.float32(0.8456), np.float32(0.7722), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7382), np.float32(0.7084), np.float32(0.6402), np.float32(0.6785), np.float32(0.0), np.float32(0.0), np.float32(0.7124), np.float32(0.6131), np.float32(0.7182), np.float32(0.5641), np.float32(0.582), np.float32(0.576), np.float32(0.8053), np.float32(0.8008), np.float32(0.5638), np.float32(0.5068), np.float32(0.0), np.float32(0.0), np.float32(0.2818), np.float32(0.5667), np.float32(0.0), np.float32(0.0), np.float32(0.7596), np.float32(0.7155), np.float32(0.8532), np.float32(0.8502)] 
+2026-10-17 01:21:13.187858: Epoch time: 220.59 s 
+2026-10-17 01:21:14.038875:  
+2026-10-17 01:21:14.039695: Epoch 101 
+2026-10-17 01:21:14.040101: Current learning rate: 0.00909 
+2026-10-17 01:25:10.967530: train_loss -0.6221 
+2026-10-17 01:25:10.968596: val_loss -0.4523 
+2026-10-17 01:25:10.969229: Pseudo dice [np.float32(0.8584), np.float32(0.4929), np.float32(0.5244), np.float32(0.6268), np.float32(0.4885), np.float32(0.5755), np.float32(0.5307), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5618), np.float32(0.522), np.float32(0.5675), np.float32(0.5157), np.float32(0.0), np.float32(0.0), np.float32(0.5692), np.float32(0.422), np.float32(0.521), np.float32(0.4121), np.float32(0.5679), np.float32(0.482), np.float32(0.3282), np.float32(0.5606), np.float32(0.3725), np.float32(0.3574), np.float32(0.0), np.float32(0.0), np.float32(0.452), np.float32(0.5554), np.float32(0.0), np.float32(0.0), np.float32(0.4788), np.float32(0.4733), np.float32(0.5508), np.float32(0.5632)] 
+2026-10-17 01:25:10.969562: Epoch time: 236.93 s 
+2026-10-17 01:25:11.834785:  
+2026-10-17 01:25:11.835489: Epoch 102 
+2026-10-17 01:25:11.835947: Current learning rate: 0.00908 
+2026-10-17 01:28:46.648272: train_loss -0.6409 
+2026-10-17 01:28:46.649080: val_loss -0.4723 
+2026-10-17 01:28:46.649854: Pseudo dice [np.float32(0.8981), np.float32(0.5814), np.float32(0.5498), np.float32(0.6773), np.float32(0.512), np.float32(0.6282), np.float32(0.5397), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6007), np.float32(0.5785), np.float32(0.5306), np.float32(0.5356), np.float32(0.0), np.float32(0.0), np.float32(0.5927), np.float32(0.4896), np.float32(0.5397), np.float32(0.4626), np.float32(0.5037), np.float32(0.472), np.float32(0.4756), np.float32(0.5271), np.float32(0.3884), np.float32(0.3944), np.float32(0.0), np.float32(0.0), np.float32(0.2373), np.float32(0.591), np.float32(0.0), np.float32(0.0), np.float32(0.5123), np.float32(0.4842), np.float32(0.5604), np.float32(0.5502)] 
+2026-10-17 01:28:46.650280: Epoch time: 214.82 s 
+2026-10-17 01:28:47.435787:  
+2026-10-17 01:28:47.437088: Epoch 103 
+2026-10-17 01:28:47.437533: Current learning rate: 0.00907 
+2026-10-17 01:32:21.801339: train_loss -0.649 
+2026-10-17 01:32:21.802179: val_loss -0.4407 
+2026-10-17 01:32:21.802900: Pseudo dice [np.float32(0.8519), np.float32(0.5219), np.float32(0.4736), np.float32(0.6096), np.float32(0.4375), np.float32(0.5775), np.float32(0.4709), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.541), np.float32(0.4901), np.float32(0.5422), np.float32(0.4729), np.float32(0.0), np.float32(0.0), np.float32(0.497), np.float32(0.378), np.float32(0.449), np.float32(0.3913), np.float32(0.324), np.float32(0.3809), np.float32(0.4497), np.float32(0.5122), np.float32(0.4773), np.float32(0.4072), np.float32(0.0), np.float32(0.0), np.float32(0.4134), np.float32(0.469), np.float32(0.0), np.float32(0.0), np.float32(0.5201), np.float32(0.4961), np.float32(0.5621), np.float32(0.5556)] 
+2026-10-17 01:32:21.803273: Epoch time: 214.37 s 
+2026-10-17 01:32:22.658682:  
+2026-10-17 01:32:22.659456: Epoch 104 
+2026-10-17 01:32:22.659849: Current learning rate: 0.00906 
+2026-10-17 01:35:46.198634: train_loss -0.6562 
+2026-10-17 01:35:46.199719: val_loss -0.4656 
+2026-10-17 01:35:46.200278: Pseudo dice [np.float32(0.8578), np.float32(0.5682), np.float32(0.5202), np.float32(0.6419), np.float32(0.4722), np.float32(0.6583), np.float32(0.5181), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5795), np.float32(0.5477), np.float32(0.568), np.float32(0.5279), np.float32(0.0), np.float32(0.0), np.float32(0.5673), np.float32(0.4722), np.float32(0.502), np.float32(0.4237), np.float32(0.4986), np.float32(0.4582), np.float32(0.385), np.float32(0.4499), np.float32(0.5178), np.float32(0.4423), np.float32(0.0), np.float32(0.0), np.float32(0.503), np.float32(0.4708), np.float32(0.0), np.float32(0.0), np.float32(0.6052), np.float32(0.5565), np.float32(0.5967), np.float32(0.6006)] 
+2026-10-17 01:35:46.200748: Epoch time: 203.54 s 
+2026-10-17 01:35:47.038311:  
+2026-10-17 01:35:47.039129: Epoch 105 
+2026-10-17 01:35:47.039529: Current learning rate: 0.00905 
+2026-10-17 01:39:25.978006: train_loss -0.6489 
+2026-10-17 01:39:25.979052: val_loss -0.4493 
+2026-10-17 01:39:25.979575: Pseudo dice [np.float32(0.8722), np.float32(0.4991), np.float32(0.4908), np.float32(0.6307), np.float32(0.4576), np.float32(0.6087), np.float32(0.4948), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5998), np.float32(0.5558), np.float32(0.587), np.float32(0.5717), np.float32(0.0), np.float32(0.0), np.float32(0.5322), np.float32(0.3983), np.float32(0.5123), np.float32(0.4679), np.float32(0.4853), np.float32(0.3624), np.float32(0.3211), np.float32(0.4101), np.float32(0.3896), np.float32(0.4143), np.float32(0.0), np.float32(0.0), np.float32(0.3303), np.float32(0.5091), np.float32(0.0), np.float32(0.0), np.float32(0.5241), np.float32(0.5046), np.float32(0.5178), np.float32(0.5388)] 
+2026-10-17 01:39:25.980008: Epoch time: 218.94 s 
+2026-10-17 01:39:26.762890:  
+2026-10-17 01:39:26.763655: Epoch 106 
+2026-10-17 01:39:26.764083: Current learning rate: 0.00904 
+2026-10-17 01:42:46.394575: train_loss -0.6547 
+2026-10-17 01:42:46.395783: val_loss -0.4812 
+2026-10-17 01:42:46.396349: Pseudo dice [np.float32(0.8751), np.float32(0.539), np.float32(0.551), np.float32(0.6853), np.float32(0.506), np.float32(0.6649), np.float32(0.5609), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6294), np.float32(0.5888), np.float32(0.6625), np.float32(0.5898), np.float32(0.0), np.float32(0.0), np.float32(0.5832), np.float32(0.47), np.float32(0.5503), np.float32(0.43), np.float32(0.4701), np.float32(0.4445), np.float32(0.3719), np.float32(0.5172), np.float32(0.3987), np.float32(0.3681), np.float32(0.0), np.float32(0.0), np.float32(0.2425), np.float32(0.5267), np.float32(0.0), np.float32(0.0), np.float32(0.5695), np.float32(0.5438), np.float32(0.5911), np.float32(0.5895)] 
+2026-10-17 01:42:46.396714: Epoch time: 199.63 s 
+2026-10-17 01:42:46.396900: Yayy! New best EMA pseudo Dice: 0.375 
+2026-10-17 01:42:49.013305:  
+2026-10-17 01:42:49.014182: Epoch 107 
+2026-10-17 01:42:49.014578: Current learning rate: 0.00903 
+2026-10-17 01:46:20.741047: train_loss -0.6547 
+2026-10-17 01:46:20.741979: val_loss -0.4481 
+2026-10-17 01:46:20.742645: Pseudo dice [np.float32(0.8925), np.float32(0.518), np.float32(0.5149), np.float32(0.6522), np.float32(0.4611), np.float32(0.6198), np.float32(0.4646), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.594), np.float32(0.5596), np.float32(0.6126), np.float32(0.5609), np.float32(0.0), np.float32(0.0), np.float32(0.5429), np.float32(0.4324), np.float32(0.493), np.float32(0.3912), np.float32(0.4106), np.float32(0.3828), np.float32(0.5034), np.float32(0.3898), np.float32(0.451), np.float32(0.332), np.float32(0.0), np.float32(0.0), np.float32(0.4631), np.float32(0.4463), np.float32(0.0), np.float32(0.0), np.float32(0.5752), np.float32(0.5284), np.float32(0.5569), np.float32(0.5553)] 
+2026-10-17 01:46:20.743127: Epoch time: 211.73 s 
+2026-10-17 01:46:20.743335: Yayy! New best EMA pseudo Dice: 0.37610000371932983 
+2026-10-17 01:46:22.166073:  
+2026-10-17 01:46:22.166757: Epoch 108 
+2026-10-17 01:46:22.167126: Current learning rate: 0.00902 
+2026-10-17 01:50:14.284452: train_loss -0.6405 
+2026-10-17 01:50:14.284948: val_loss -0.5506 
+2026-10-17 01:50:14.285275: Pseudo dice [np.float32(0.7986), np.float32(0.6419), np.float32(0.5946), np.float32(0.747), np.float32(0.6395), np.float32(0.7922), np.float32(0.7035), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6364), np.float32(0.6217), np.float32(0.5987), np.float32(0.6191), np.float32(0.0), np.float32(0.0), np.float32(0.6642), np.float32(0.543), np.float32(0.6399), np.float32(0.5193), np.float32(0.5214), np.float32(0.5283), np.float32(0.724), np.float32(0.7312), np.float32(0.4868), np.float32(0.4568), np.float32(0.0), np.float32(0.0), np.float32(0.4715), np.float32(0.6119), np.float32(0.0), np.float32(0.0), np.float32(0.6705), np.float32(0.6594), np.float32(0.7805), np.float32(0.801)] 
+2026-10-17 01:50:14.286203: Epoch time: 232.12 s 
+2026-10-17 01:50:14.286556: Yayy! New best EMA pseudo Dice: 0.3862999975681305 
+2026-10-17 01:50:19.169718:  
+2026-10-17 01:50:19.170389: Epoch 109 
+2026-10-17 01:50:19.170802: Current learning rate: 0.00901 
+2026-10-17 01:53:56.280858: train_loss -0.6317 
+2026-10-17 01:53:56.281574: val_loss -0.4764 
+2026-10-17 01:53:56.282448: Pseudo dice [np.float32(0.8531), np.float32(0.5353), np.float32(0.502), np.float32(0.5956), np.float32(0.4815), np.float32(0.6096), np.float32(0.5219), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5332), np.float32(0.5238), np.float32(0.5376), np.float32(0.5514), np.float32(0.0), np.float32(0.0), np.float32(0.5089), np.float32(0.4257), np.float32(0.464), np.float32(0.4007), np.float32(0.3603), np.float32(0.3262), np.float32(0.6505), np.float32(0.6515), np.float32(0.3858), np.float32(0.397), np.float32(0.0), np.float32(0.0), np.float32(0.1891), np.float32(0.5056), np.float32(0.0), np.float32(0.0), np.float32(0.409), np.float32(0.39), np.float32(0.5771), np.float32(0.5785)] 
+2026-10-17 01:53:56.282958: Epoch time: 217.11 s 
+2026-10-17 01:53:57.097551:  
+2026-10-17 01:53:57.098173: Epoch 110 
+2026-10-17 01:53:57.098611: Current learning rate: 0.009 
+2026-10-17 01:57:51.750978: train_loss -0.6371 
+2026-10-17 01:57:51.751879: val_loss -0.4434 
+2026-10-17 01:57:51.752448: Pseudo dice [np.float32(0.8543), np.float32(0.4864), np.float32(0.4922), np.float32(0.63), np.float32(0.4753), np.float32(0.5851), np.float32(0.4708), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5621), np.float32(0.5276), np.float32(0.565), np.float32(0.514), np.float32(0.0), np.float32(0.0), np.float32(0.559), np.float32(0.41), np.float32(0.4845), np.float32(0.3435), np.float32(0.3736), np.float32(0.3217), np.float32(0.3621), np.float32(0.5142), np.float32(0.4786), np.float32(0.3987), np.float32(0.0), np.float32(0.0), np.float32(0.2926), np.float32(0.576), np.float32(0.0), np.float32(0.0), np.float32(0.496), np.float32(0.4253), np.float32(0.5089), np.float32(0.4427)] 
+2026-10-17 01:57:51.752751: Epoch time: 234.66 s 
+2026-10-17 01:57:52.557426:  
+2026-10-17 01:57:52.558254: Epoch 111 
+2026-10-17 01:57:52.558780: Current learning rate: 0.009 
+2026-10-17 02:02:00.614717: train_loss -0.6418 
+2026-10-17 02:02:00.615333: val_loss -0.4194 
+2026-10-17 02:02:00.616026: Pseudo dice [np.float32(0.8166), np.float32(0.4202), np.float32(0.3641), np.float32(0.55), np.float32(0.3768), np.float32(0.4847), np.float32(0.3916), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4861), np.float32(0.4548), np.float32(0.4563), np.float32(0.4419), np.float32(0.0), np.float32(0.0), np.float32(0.3881), np.float32(0.2405), np.float32(0.3401), np.float32(0.2986), np.float32(0.3163), np.float32(0.269), np.float32(0.5827), np.float32(0.6144), np.float32(0.302), np.float32(0.4235), np.float32(0.0), np.float32(0.0), np.float32(0.166), np.float32(0.3259), np.float32(0.0), np.float32(0.0), np.float32(0.395), np.float32(0.2941), np.float32(0.4865), np.float32(0.4178)] 
+2026-10-17 02:02:00.616864: Epoch time: 248.06 s 
+2026-10-17 02:02:01.531595:  
+2026-10-17 02:02:01.532340: Epoch 112 
+2026-10-17 02:02:01.532861: Current learning rate: 0.00899 
+2026-10-17 02:05:35.570858: train_loss -0.6536 
+2026-10-17 02:05:35.571517: val_loss -0.4192 
+2026-10-17 02:05:35.572490: Pseudo dice [np.float32(0.8713), np.float32(0.4089), np.float32(0.3763), np.float32(0.4699), np.float32(0.3452), np.float32(0.4439), np.float32(0.3946), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4262), np.float32(0.3708), np.float32(0.4565), np.float32(0.4017), np.float32(0.0), np.float32(0.0), np.float32(0.4538), np.float32(0.3519), np.float32(0.3652), np.float32(0.3119), np.float32(0.2004), np.float32(0.2535), np.float32(0.3654), np.float32(0.4956), np.float32(0.3393), np.float32(0.3955), np.float32(0.0), np.float32(0.0), np.float32(0.0442), np.float32(0.4551), np.float32(0.0), np.float32(0.0), np.float32(0.2844), np.float32(0.3621), np.float32(0.3528), np.float32(0.375)] 
+2026-10-17 02:05:35.573087: Epoch time: 214.04 s 
+2026-10-17 02:05:36.384781:  
+2026-10-17 02:05:36.385431: Epoch 113 
+2026-10-17 02:05:36.385827: Current learning rate: 0.00898 
+2026-10-17 02:09:35.479806: train_loss -0.6607 
+2026-10-17 02:09:35.480356: val_loss -0.3355 
+2026-10-17 02:09:35.481114: Pseudo dice [np.float32(0.8245), np.float32(0.2855), np.float32(0.284), np.float32(0.4064), np.float32(0.2659), np.float32(0.296), np.float32(0.2537), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.296), np.float32(0.2382), np.float32(0.3147), np.float32(0.2367), np.float32(0.0), np.float32(0.0), np.float32(0.3177), np.float32(0.2072), np.float32(0.1858), np.float32(0.1432), np.float32(0.0745), np.float32(0.1911), np.float32(0.1716), np.float32(0.4204), np.float32(0.2826), np.float32(0.1665), np.float32(0.0), np.float32(0.0), np.float32(0.0312), np.float32(0.3072), np.float32(0.0), np.float32(0.0), np.float32(0.2069), np.float32(0.1708), np.float32(0.2708), np.float32(0.2443)] 
+2026-10-17 02:09:35.481417: Epoch time: 239.1 s 
+2026-10-17 02:09:36.249477:  
+2026-10-17 02:09:36.250146: Epoch 114 
+2026-10-17 02:09:36.250359: Current learning rate: 0.00897 
+2026-10-17 02:13:30.266912: train_loss -0.6442 
+2026-10-17 02:13:30.267917: val_loss -0.2682 
+2026-10-17 02:13:30.268648: Pseudo dice [np.float32(0.854), np.float32(0.1822), np.float32(0.1795), np.float32(0.1772), np.float32(0.1207), np.float32(0.2102), np.float32(0.1866), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2472), np.float32(0.2067), np.float32(0.3059), np.float32(0.1955), np.float32(0.0), np.float32(0.0), np.float32(0.1895), np.float32(0.1401), np.float32(0.1792), np.float32(0.2115), np.float32(0.0947), np.float32(0.1705), np.float32(0.2312), np.float32(0.2368), np.float32(0.2592), np.float32(0.2695), np.float32(0.0), np.float32(0.0), np.float32(0.0323), np.float32(0.3521), np.float32(0.0), np.float32(0.0), np.float32(0.1276), np.float32(0.0922), np.float32(0.1503), np.float32(0.2386)] 
+2026-10-17 02:13:30.268904: Epoch time: 234.02 s 
+2026-10-17 02:13:31.085789:  
+2026-10-17 02:13:31.086673: Epoch 115 
+2026-10-17 02:13:31.087214: Current learning rate: 0.00896 
+2026-10-17 02:17:09.544796: train_loss -0.6491 
+2026-10-17 02:17:09.545378: val_loss -0.5157 
+2026-10-17 02:17:09.545861: Pseudo dice [np.float32(0.8644), np.float32(0.6345), np.float32(0.5866), np.float32(0.682), np.float32(0.5788), np.float32(0.6516), np.float32(0.5845), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5728), np.float32(0.5888), np.float32(0.5431), np.float32(0.5718), np.float32(0.0), np.float32(0.0), np.float32(0.6271), np.float32(0.5006), np.float32(0.555), np.float32(0.4687), np.float32(0.4627), np.float32(0.4542), np.float32(0.6759), np.float32(0.6841), np.float32(0.4583), np.float32(0.4584), np.float32(0.0), np.float32(0.0), np.float32(0.215), np.float32(0.4944), np.float32(0.0), np.float32(0.0), np.float32(0.5173), np.float32(0.3133), np.float32(0.6711), np.float32(0.6477)] 
+2026-10-17 02:17:09.546557: Epoch time: 218.46 s 
+2026-10-17 02:17:10.358113:  
+2026-10-17 02:17:10.358829: Epoch 116 
+2026-10-17 02:17:10.359190: Current learning rate: 0.00895 
+2026-10-17 02:20:58.928830: train_loss -0.6607 
+2026-10-17 02:20:58.929795: val_loss -0.378 
+2026-10-17 02:20:58.930391: Pseudo dice [np.float32(0.8128), np.float32(0.4285), np.float32(0.374), np.float32(0.5044), np.float32(0.3846), np.float32(0.4784), np.float32(0.414), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4173), np.float32(0.438), np.float32(0.353), np.float32(0.333), np.float32(0.0), np.float32(0.0), np.float32(0.4226), np.float32(0.2986), np.float32(0.3105), np.float32(0.2777), np.float32(0.1903), np.float32(0.2524), np.float32(0.4452), np.float32(0.4471), np.float32(0.3375), np.float32(0.2925), np.float32(0.0), np.float32(0.0), np.float32(0.1182), np.float32(0.4437), np.float32(0.0), np.float32(0.0), np.float32(0.3722), np.float32(0.3841), np.float32(0.4143), np.float32(0.379)] 
+2026-10-17 02:20:58.931068: Epoch time: 228.57 s 
+2026-10-17 02:20:59.779620:  
+2026-10-17 02:20:59.780301: Epoch 117 
+2026-10-17 02:20:59.780625: Current learning rate: 0.00894 
+2026-10-17 02:24:37.013805: train_loss -0.6494 
+2026-10-17 02:24:37.014665: val_loss -0.6088 
+2026-10-17 02:24:37.015137: Pseudo dice [np.float32(0.8044), np.float32(0.8264), np.float32(0.732), np.float32(0.8631), np.float32(0.7074), np.float32(0.8486), np.float32(0.7805), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7216), np.float32(0.7117), np.float32(0.6628), np.float32(0.7243), np.float32(0.0), np.float32(0.0), np.float32(0.7277), np.float32(0.5605), np.float32(0.6879), np.float32(0.5624), np.float32(0.6584), np.float32(0.6045), np.float32(0.7129), np.float32(0.7744), np.float32(0.512), np.float32(0.5546), np.float32(0.0), np.float32(0.0), np.float32(0.4933), np.float32(0.4501), np.float32(0.0), np.float32(0.0), np.float32(0.7819), np.float32(0.7306), np.float32(0.874), np.float32(0.8384)] 
+2026-10-17 02:24:37.015649: Epoch time: 217.24 s 
+2026-10-17 02:24:37.813308:  
+2026-10-17 02:24:37.814123: Epoch 118 
+2026-10-17 02:24:37.814495: Current learning rate: 0.00893 
+2026-10-17 02:28:03.130947: train_loss -0.6487 
+2026-10-17 02:28:03.131681: val_loss -0.5491 
+2026-10-17 02:28:03.132504: Pseudo dice [np.float32(0.8758), np.float32(0.6684), np.float32(0.6528), np.float32(0.7555), np.float32(0.6227), np.float32(0.7153), np.float32(0.6639), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6425), np.float32(0.635), np.float32(0.5733), np.float32(0.5707), np.float32(0.0), np.float32(0.0), np.float32(0.638), np.float32(0.5121), np.float32(0.6103), np.float32(0.53), np.float32(0.6191), np.float32(0.5143), np.float32(0.5357), np.float32(0.6403), np.float32(0.5323), np.float32(0.4319), np.float32(0.0), np.float32(0.0), np.float32(0.3945), np.float32(0.615), np.float32(0.0), np.float32(0.0), np.float32(0.5817), np.float32(0.4851), np.float32(0.7009), np.float32(0.7042)] 
+2026-10-17 02:28:03.132886: Epoch time: 205.32 s 
+2026-10-17 02:28:03.985388:  
+2026-10-17 02:28:03.986000: Epoch 119 
+2026-10-17 02:28:03.986286: Current learning rate: 0.00892 
+2026-10-17 02:31:33.882174: train_loss -0.6358 
+2026-10-17 02:31:33.882701: val_loss -0.4371 
+2026-10-17 02:31:33.883285: Pseudo dice [np.float32(0.8011), np.float32(0.5011), np.float32(0.5075), np.float32(0.5408), np.float32(0.389), np.float32(0.4967), np.float32(0.3872), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5062), np.float32(0.481), np.float32(0.591), np.float32(0.5307), np.float32(0.0), np.float32(0.0), np.float32(0.4806), np.float32(0.3941), np.float32(0.376), np.float32(0.289), np.float32(0.4748), np.float32(0.3608), np.float32(0.432), np.float32(0.5825), np.float32(0.3179), np.float32(0.3578), np.float32(0.0), np.float32(0.0), np.float32(0.385), np.float32(0.4974), np.float32(0.0), np.float32(0.0), np.float32(0.4109), np.float32(0.2681), np.float32(0.51), np.float32(0.543)] 
+2026-10-17 02:31:33.884355: Epoch time: 209.9 s 
+2026-10-17 02:31:34.668753:  
+2026-10-17 02:31:34.669477: Epoch 120 
+2026-10-17 02:31:34.669848: Current learning rate: 0.00891 
+2026-10-17 02:35:15.041300: train_loss -0.6286 
+2026-10-17 02:35:15.042020: val_loss -0.4975 
+2026-10-17 02:35:15.042949: Pseudo dice [np.float32(0.8488), np.float32(0.6138), np.float32(0.5774), np.float32(0.6995), np.float32(0.544), np.float32(0.6736), np.float32(0.5485), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6043), np.float32(0.5524), np.float32(0.5498), np.float32(0.5302), np.float32(0.0), np.float32(0.0), np.float32(0.5561), np.float32(0.3964), np.float32(0.554), np.float32(0.4443), np.float32(0.5088), np.float32(0.3862), np.float32(0.7182), np.float32(0.6824), np.float32(0.5716), np.float32(0.4267), np.float32(0.0), np.float32(0.0), np.float32(0.3828), np.float32(0.3676), np.float32(0.0), np.float32(0.0), np.float32(0.6441), np.float32(0.5771), np.float32(0.7334), np.float32(0.6997)] 
+2026-10-17 02:35:15.043410: Epoch time: 220.37 s 
+2026-10-17 02:35:15.935314:  
+2026-10-17 02:35:15.936230: Epoch 121 
+2026-10-17 02:35:15.936653: Current learning rate: 0.0089 
+2026-10-17 02:38:52.784354: train_loss -0.6248 
+2026-10-17 02:38:52.784970: val_loss -0.4969 
+2026-10-17 02:38:52.785477: Pseudo dice [np.float32(0.8858), np.float32(0.5827), np.float32(0.5391), np.float32(0.694), np.float32(0.5594), np.float32(0.6669), np.float32(0.5966), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5549), np.float32(0.5406), np.float32(0.5057), np.float32(0.4833), np.float32(0.0), np.float32(0.0), np.float32(0.5581), np.float32(0.4492), np.float32(0.5514), np.float32(0.4494), np.float32(0.4989), np.float32(0.4112), np.float32(0.6358), np.float32(0.6517), np.float32(0.4537), np.float32(0.3878), np.float32(0.0), np.float32(0.0), np.float32(0.3332), np.float32(0.5462), np.float32(0.0), np.float32(0.0), np.float32(0.5753), np.float32(0.5926), np.float32(0.6523), np.float32(0.6065)] 
+2026-10-17 02:38:52.785829: Epoch time: 216.85 s 
+2026-10-17 02:38:53.593271:  
+2026-10-17 02:38:53.593705: Epoch 122 
+2026-10-17 02:38:53.594031: Current learning rate: 0.00889 
+2026-10-17 02:42:37.578919: train_loss -0.6337 
+2026-10-17 02:42:37.579821: val_loss -0.5267 
+2026-10-17 02:42:37.580679: Pseudo dice [np.float32(0.857), np.float32(0.6254), np.float32(0.5628), np.float32(0.7129), np.float32(0.5701), np.float32(0.698), np.float32(0.621), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6238), np.float32(0.6327), np.float32(0.5526), np.float32(0.5599), np.float32(0.0), np.float32(0.0), np.float32(0.6159), np.float32(0.4602), np.float32(0.553), np.float32(0.4697), np.float32(0.5713), np.float32(0.4489), np.float32(0.6685), np.float32(0.6892), np.float32(0.4819), np.float32(0.505), np.float32(0.0), np.float32(0.0), np.float32(0.3148), np.float32(0.3825), np.float32(0.0), np.float32(0.0), np.float32(0.555), np.float32(0.4833), np.float32(0.6574), np.float32(0.6736)] 
+2026-10-17 02:42:37.581209: Epoch time: 223.99 s 
+2026-10-17 02:42:38.476709:  
+2026-10-17 02:42:38.477487: Epoch 123 
+2026-10-17 02:42:38.477819: Current learning rate: 0.00889 
+2026-10-17 02:46:25.929413: train_loss -0.6186 
+2026-10-17 02:46:25.930299: val_loss -0.4935 
+2026-10-17 02:46:25.930868: Pseudo dice [np.float32(0.829), np.float32(0.562), np.float32(0.5704), np.float32(0.6615), np.float32(0.4962), np.float32(0.6104), np.float32(0.5749), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5343), np.float32(0.516), np.float32(0.4736), np.float32(0.4454), np.float32(0.0), np.float32(0.0), np.float32(0.4885), np.float32(0.3565), np.float32(0.5059), np.float32(0.4506), np.float32(0.4944), np.float32(0.4276), np.float32(0.5083), np.float32(0.601), np.float32(0.3722), np.float32(0.4851), np.float32(0.0), np.float32(0.0), np.float32(0.4549), np.float32(0.4123), np.float32(0.0), np.float32(0.0), np.float32(0.4833), np.float32(0.4886), np.float32(0.5989), np.float32(0.6153)] 
+2026-10-17 02:46:25.931238: Epoch time: 227.45 s 
+2026-10-17 02:46:26.723401:  
+2026-10-17 02:46:26.724128: Epoch 124 
+2026-10-17 02:46:26.724413: Current learning rate: 0.00888 
+2026-10-17 02:49:43.895333: train_loss -0.6399 
+2026-10-17 02:49:43.896630: val_loss -0.5298 
+2026-10-17 02:49:43.897377: Pseudo dice [np.float32(0.8831), np.float32(0.6082), np.float32(0.5734), np.float32(0.7248), np.float32(0.5713), np.float32(0.7178), np.float32(0.6238), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.611), np.float32(0.6052), np.float32(0.6161), np.float32(0.5964), np.float32(0.0), np.float32(0.0), np.float32(0.6262), np.float32(0.5267), np.float32(0.5805), np.float32(0.465), np.float32(0.5907), np.float32(0.5561), np.float32(0.6164), np.float32(0.681), np.float32(0.3657), np.float32(0.5719), np.float32(0.0), np.float32(0.0), np.float32(0.3381), np.float32(0.5644), np.float32(0.0), np.float32(0.0), np.float32(0.6336), np.float32(0.5886), np.float32(0.7057), np.float32(0.711)] 
+2026-10-17 02:49:43.898007: Epoch time: 197.17 s 
+2026-10-17 02:49:43.898366: Yayy! New best EMA pseudo Dice: 0.3871999979019165 
+2026-10-17 02:49:45.239341:  
+2026-10-17 02:49:45.240070: Epoch 125 
+2026-10-17 02:49:45.240432: Current learning rate: 0.00887 
+2026-10-17 02:53:28.744804: train_loss -0.6616 
+2026-10-17 02:53:28.745433: val_loss -0.4013 
+2026-10-17 02:53:28.745976: Pseudo dice [np.float32(0.867), np.float32(0.3715), np.float32(0.3729), np.float32(0.4507), np.float32(0.356), np.float32(0.4176), np.float32(0.3521), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3667), np.float32(0.3682), np.float32(0.3666), np.float32(0.3877), np.float32(0.0), np.float32(0.0), np.float32(0.3805), np.float32(0.2925), np.float32(0.3161), np.float32(0.2591), np.float32(0.1999), np.float32(0.2753), np.float32(0.459), np.float32(0.577), np.float32(0.225), np.float32(0.386), np.float32(0.0), np.float32(0.0), np.float32(0.104), np.float32(0.3969), np.float32(0.0), np.float32(0.0), np.float32(0.2631), np.float32(0.2388), np.float32(0.3644), np.float32(0.4064)] 
+2026-10-17 02:53:28.746445: Epoch time: 223.51 s 
+2026-10-17 02:53:29.618561:  
+2026-10-17 02:53:29.619420: Epoch 126 
+2026-10-17 02:53:29.619993: Current learning rate: 0.00886 
+2026-10-17 02:57:07.620871: train_loss -0.6624 
+2026-10-17 02:57:07.621648: val_loss -0.5306 
+2026-10-17 02:57:07.622878: Pseudo dice [np.float32(0.8555), np.float32(0.6482), np.float32(0.604), np.float32(0.7443), np.float32(0.5907), np.float32(0.7447), np.float32(0.6483), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5772), np.float32(0.5876), np.float32(0.533), np.float32(0.5496), np.float32(0.0), np.float32(0.0), np.float32(0.5683), np.float32(0.4179), np.float32(0.5832), np.float32(0.5117), np.float32(0.4831), np.float32(0.3755), np.float32(0.6978), np.float32(0.7618), np.float32(0.5699), np.float32(0.5014), np.float32(0.0), np.float32(0.0), np.float32(0.3655), np.float32(0.5406), np.float32(0.0), np.float32(0.0), np.float32(0.6135), np.float32(0.6136), np.float32(0.7395), np.float32(0.766)] 
+2026-10-17 02:57:07.623486: Epoch time: 218.0 s 
+2026-10-17 02:57:08.722071:  
+2026-10-17 02:57:08.722812: Epoch 127 
+2026-10-17 02:57:08.723273: Current learning rate: 0.00885 
+2026-10-17 03:00:46.294297: train_loss -0.6425 
+2026-10-17 03:00:46.295058: val_loss -0.3728 
+2026-10-17 03:00:46.295636: Pseudo dice [np.float32(0.8794), np.float32(0.3254), np.float32(0.3326), np.float32(0.4179), np.float32(0.3619), np.float32(0.3551), np.float32(0.3079), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2583), np.float32(0.2767), np.float32(0.169), np.float32(0.2914), np.float32(0.0), np.float32(0.0), np.float32(0.2924), np.float32(0.2078), np.float32(0.2171), np.float32(0.2215), np.float32(0.178), np.float32(0.1668), np.float32(0.5944), np.float32(0.4514), np.float32(0.1251), np.float32(0.2176), np.float32(0.0), np.float32(0.0), np.float32(0.0637), np.float32(0.2816), np.float32(0.0), np.float32(0.0), np.float32(0.2853), np.float32(0.3435), np.float32(0.4834), np.float32(0.3384)] 
+2026-10-17 03:00:46.295896: Epoch time: 217.57 s 
+2026-10-17 03:00:47.164528:  
+2026-10-17 03:00:47.165387: Epoch 128 
+2026-10-17 03:00:47.165764: Current learning rate: 0.00884 
+2026-10-17 03:04:17.032292: train_loss -0.6406 
+2026-10-17 03:04:17.033120: val_loss -0.4361 
+2026-10-17 03:04:17.033813: Pseudo dice [np.float32(0.8348), np.float32(0.3723), np.float32(0.4233), np.float32(0.5582), np.float32(0.3999), np.float32(0.5464), np.float32(0.4054), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5141), np.float32(0.4736), np.float32(0.5433), np.float32(0.5482), np.float32(0.0), np.float32(0.0), np.float32(0.4419), np.float32(0.35), np.float32(0.3753), np.float32(0.3008), np.float32(0.2911), np.float32(0.3047), np.float32(0.2038), np.float32(0.2678), np.float32(0.237), np.float32(0.2279), np.float32(0.0), np.float32(0.0), np.float32(0.3225), np.float32(0.3899), np.float32(0.0), np.float32(0.0), np.float32(0.4823), np.float32(0.4275), np.float32(0.4239), np.float32(0.3758)] 
+2026-10-17 03:04:17.034548: Epoch time: 209.87 s 
+2026-10-17 03:04:17.913824:  
+2026-10-17 03:04:17.914393: Epoch 129 
+2026-10-17 03:04:17.914900: Current learning rate: 0.00883 
+2026-10-17 03:08:09.065323: train_loss -0.6569 
+2026-10-17 03:08:09.065913: val_loss -0.4221 
+2026-10-17 03:08:09.067663: Pseudo dice [np.float32(0.8663), np.float32(0.4343), np.float32(0.3981), np.float32(0.537), np.float32(0.4299), np.float32(0.4687), np.float32(0.4528), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4195), np.float32(0.3932), np.float32(0.4425), np.float32(0.4819), np.float32(0.0), np.float32(0.0), np.float32(0.476), np.float32(0.3807), np.float32(0.3481), np.float32(0.3609), np.float32(0.3331), np.float32(0.3031), np.float32(0.454), np.float32(0.4407), np.float32(0.3618), np.float32(0.4259), np.float32(0.0), np.float32(0.0), np.float32(0.1186), np.float32(0.4059), np.float32(0.0), np.float32(0.0), np.float32(0.4298), np.float32(0.4484), np.float32(0.485), np.float32(0.5007)] 
+2026-10-17 03:08:09.068911: Epoch time: 231.15 s 
+2026-10-17 03:08:09.963001:  
+2026-10-17 03:08:09.963792: Epoch 130 
+2026-10-17 03:08:09.964156: Current learning rate: 0.00882 
+2026-10-17 03:11:56.811068: train_loss -0.6711 
+2026-10-17 03:11:56.811720: val_loss -0.4645 
+2026-10-17 03:11:56.812480: Pseudo dice [np.float32(0.7904), np.float32(0.5173), np.float32(0.4766), np.float32(0.5913), np.float32(0.4831), np.float32(0.5873), np.float32(0.5158), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4431), np.float32(0.48), np.float32(0.4408), np.float32(0.4099), np.float32(0.0), np.float32(0.0), np.float32(0.475), np.float32(0.4145), np.float32(0.3902), np.float32(0.3927), np.float32(0.3712), np.float32(0.3005), np.float32(0.4889), np.float32(0.5868), np.float32(0.3323), np.float32(0.4969), np.float32(0.0), np.float32(0.0), np.float32(0.1773), np.float32(0.3975), np.float32(0.0), np.float32(0.0), np.float32(0.4322), np.float32(0.4673), np.float32(0.604), np.float32(0.6062)] 
+2026-10-17 03:11:56.813209: Epoch time: 226.85 s 
+2026-10-17 03:11:57.792776:  
+2026-10-17 03:11:57.793548: Epoch 131 
+2026-10-17 03:11:57.794046: Current learning rate: 0.00881 
+2026-10-17 03:15:31.053376: train_loss -0.685 
+2026-10-17 03:15:31.054046: val_loss -0.4324 
+2026-10-17 03:15:31.054552: Pseudo dice [np.float32(0.84), np.float32(0.4605), np.float32(0.4153), np.float32(0.5035), np.float32(0.4374), np.float32(0.5021), np.float32(0.4414), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4114), np.float32(0.4105), np.float32(0.4291), np.float32(0.3678), np.float32(0.0), np.float32(0.0), np.float32(0.4542), np.float32(0.4242), np.float32(0.3358), np.float32(0.3719), np.float32(0.3239), np.float32(0.2713), np.float32(0.6147), np.float32(0.5379), np.float32(0.2925), np.float32(0.41), np.float32(0.0), np.float32(0.0), np.float32(0.1449), np.float32(0.3621), np.float32(0.0), np.float32(0.0), np.float32(0.4054), np.float32(0.3928), np.float32(0.5142), np.float32(0.5255)] 
+2026-10-17 03:15:31.055151: Epoch time: 213.26 s 
+2026-10-17 03:15:31.906868:  
+2026-10-17 03:15:31.907607: Epoch 132 
+2026-10-17 03:15:31.908105: Current learning rate: 0.0088 
+2026-10-17 03:19:16.558118: train_loss -0.6777 
+2026-10-17 03:19:16.558878: val_loss -0.3952 
+2026-10-17 03:19:16.559971: Pseudo dice [np.float32(0.8541), np.float32(0.3666), np.float32(0.3483), np.float32(0.4672), np.float32(0.3625), np.float32(0.4071), np.float32(0.3551), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3629), np.float32(0.3382), np.float32(0.4176), np.float32(0.3993), np.float32(0.0), np.float32(0.0), np.float32(0.4165), np.float32(0.3335), np.float32(0.3128), np.float32(0.2884), np.float32(0.2395), np.float32(0.3085), np.float32(0.2912), np.float32(0.458), np.float32(0.2565), np.float32(0.3023), np.float32(0.0), np.float32(0.0), np.float32(0.0517), np.float32(0.3955), np.float32(0.0), np.float32(0.0), np.float32(0.3767), np.float32(0.333), np.float32(0.3979), np.float32(0.3607)] 
+2026-10-17 03:19:16.560544: Epoch time: 224.65 s 
+2026-10-17 03:19:17.387802:  
+2026-10-17 03:19:17.388607: Epoch 133 
+2026-10-17 03:19:17.389079: Current learning rate: 0.00879 
+2026-10-17 03:23:01.182526: train_loss -0.6758 
+2026-10-17 03:23:01.183228: val_loss -0.4927 
+2026-10-17 03:23:01.183869: Pseudo dice [np.float32(0.8487), np.float32(0.5898), np.float32(0.489), np.float32(0.5842), np.float32(0.5451), np.float32(0.5495), np.float32(0.5339), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4259), np.float32(0.4293), np.float32(0.4937), np.float32(0.5636), np.float32(0.0), np.float32(0.0), np.float32(0.5558), np.float32(0.4332), np.float32(0.5232), np.float32(0.5135), np.float32(0.3816), np.float32(0.327), np.float32(0.643), np.float32(0.674), np.float32(0.3331), np.float32(0.4637), np.float32(0.0), np.float32(0.0), np.float32(0.0641), np.float32(0.4312), np.float32(0.0), np.float32(0.0), np.float32(0.4249), np.float32(0.4088), np.float32(0.5359), np.float32(0.5586)] 
+2026-10-17 03:23:01.184277: Epoch time: 223.8 s 
+2026-10-17 03:23:02.010221:  
+2026-10-17 03:23:02.010903: Epoch 134 
+2026-10-17 03:23:02.011320: Current learning rate: 0.00879 
+2026-10-17 03:27:13.584791: train_loss -0.6675 
+2026-10-17 03:27:13.585604: val_loss -0.5536 
+2026-10-17 03:27:13.586428: Pseudo dice [np.float32(0.8541), np.float32(0.6369), np.float32(0.5924), np.float32(0.6886), np.float32(0.6188), np.float32(0.6776), np.float32(0.6329), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5671), np.float32(0.5762), np.float32(0.5426), np.float32(0.609), np.float32(0.0), np.float32(0.0), np.float32(0.6268), np.float32(0.5085), np.float32(0.5713), np.float32(0.5208), np.float32(0.4457), np.float32(0.3429), np.float32(0.7495), np.float32(0.7608), np.float32(0.4055), np.float32(0.5539), np.float32(0.0), np.float32(0.0), np.float32(0.1633), np.float32(0.434), np.float32(0.0), np.float32(0.0), np.float32(0.555), np.float32(0.5453), np.float32(0.6544), np.float32(0.6612)] 
+2026-10-17 03:27:13.586927: Epoch time: 251.58 s 
+2026-10-17 03:27:14.509473:  
+2026-10-17 03:27:14.509949: Epoch 135 
+2026-10-17 03:27:14.510287: Current learning rate: 0.00878 
+2026-10-17 03:30:59.309264: train_loss -0.6737 
+2026-10-17 03:30:59.310285: val_loss -0.5191 
+2026-10-17 03:30:59.310851: Pseudo dice [np.float32(0.875), np.float32(0.6221), np.float32(0.5949), np.float32(0.7116), np.float32(0.5841), np.float32(0.7035), np.float32(0.5981), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5535), np.float32(0.5811), np.float32(0.5397), np.float32(0.5217), np.float32(0.0), np.float32(0.0), np.float32(0.6095), np.float32(0.5191), np.float32(0.5629), np.float32(0.5099), np.float32(0.4458), np.float32(0.3986), np.float32(0.6808), np.float32(0.5979), np.float32(0.5085), np.float32(0.5096), np.float32(0.0), np.float32(0.0), np.float32(0.5048), np.float32(0.4262), np.float32(0.0), np.float32(0.0), np.float32(0.5992), np.float32(0.5974), np.float32(0.7259), np.float32(0.7446)] 
+2026-10-17 03:30:59.311269: Epoch time: 224.8 s 
+2026-10-17 03:31:00.192429:  
+2026-10-17 03:31:00.193211: Epoch 136 
+2026-10-17 03:31:00.193557: Current learning rate: 0.00877 
+2026-10-17 03:34:38.290646: train_loss -0.664 
+2026-10-17 03:34:38.291124: val_loss -0.4359 
+2026-10-17 03:34:38.291599: Pseudo dice [np.float32(0.8728), np.float32(0.4731), np.float32(0.4073), np.float32(0.5774), np.float32(0.4875), np.float32(0.5512), np.float32(0.4862), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4269), np.float32(0.4447), np.float32(0.4099), np.float32(0.4108), np.float32(0.0), np.float32(0.0), np.float32(0.5077), np.float32(0.4214), np.float32(0.4337), np.float32(0.4133), np.float32(0.3887), np.float32(0.3118), np.float32(0.6363), np.float32(0.5833), np.float32(0.3324), np.float32(0.3632), np.float32(0.0), np.float32(0.0), np.float32(0.1907), np.float32(0.4609), np.float32(0.0), np.float32(0.0), np.float32(0.4897), np.float32(0.4721), np.float32(0.57), np.float32(0.5708)] 
+2026-10-17 03:34:38.292417: Epoch time: 218.1 s 
+2026-10-17 03:34:39.140695:  
+2026-10-17 03:34:39.141366: Epoch 137 
+2026-10-17 03:34:39.141712: Current learning rate: 0.00876 
+2026-10-17 03:38:56.933862: train_loss -0.6052 
+2026-10-17 03:38:56.934653: val_loss -0.5319 
+2026-10-17 03:38:56.935331: Pseudo dice [np.float32(0.8862), np.float32(0.6372), np.float32(0.6451), np.float32(0.714), np.float32(0.5716), np.float32(0.6754), np.float32(0.6243), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6416), np.float32(0.6042), np.float32(0.6192), np.float32(0.6365), np.float32(0.0), np.float32(0.0), np.float32(0.6441), np.float32(0.5421), np.float32(0.5994), np.float32(0.5357), np.float32(0.5876), np.float32(0.4869), np.float32(0.6782), np.float32(0.6904), np.float32(0.3218), np.float32(0.2667), np.float32(0.0), np.float32(0.0), np.float32(0.1659), np.float32(0.4605), np.float32(0.0), np.float32(0.0), np.float32(0.6129), np.float32(0.5436), np.float32(0.7024), np.float32(0.6835)] 
+2026-10-17 03:38:56.935825: Epoch time: 257.8 s 
+2026-10-17 03:38:57.807643:  
+2026-10-17 03:38:57.808357: Epoch 138 
+2026-10-17 03:38:57.808702: Current learning rate: 0.00875 
+2026-10-17 03:42:42.032302: train_loss -0.592 
+2026-10-17 03:42:42.033001: val_loss -0.3215 
+2026-10-17 03:42:42.034055: Pseudo dice [np.float32(0.8478), np.float32(0.1689), np.float32(0.2473), np.float32(0.2838), np.float32(0.2163), np.float32(0.1838), np.float32(0.2742), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1917), np.float32(0.1849), np.float32(0.3812), np.float32(0.3447), np.float32(0.0), np.float32(0.0), np.float32(0.214), np.float32(0.1332), np.float32(0.3136), np.float32(0.228), np.float32(0.2721), np.float32(0.251), np.float32(0.2008), np.float32(0.3368), np.float32(0.1383), np.float32(0.0647), np.float32(0.0), np.float32(0.0), np.float32(0.1264), np.float32(0.3002), np.float32(0.0), np.float32(0.0), np.float32(0.2472), np.float32(0.1977), np.float32(0.2634), np.float32(0.2288)] 
+2026-10-17 03:42:42.034475: Epoch time: 224.23 s 
+2026-10-17 03:42:42.930642:  
+2026-10-17 03:42:42.931484: Epoch 139 
+2026-10-17 03:42:42.931786: Current learning rate: 0.00874 
+2026-10-17 03:46:30.918249: train_loss -0.6289 
+2026-10-17 03:46:30.918855: val_loss -0.5988 
+2026-10-17 03:46:30.919197: Pseudo dice [np.float32(0.8637), np.float32(0.7717), np.float32(0.6827), np.float32(0.8483), np.float32(0.7176), np.float32(0.848), np.float32(0.76), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6927), np.float32(0.6965), np.float32(0.639), np.float32(0.6904), np.float32(0.0), np.float32(0.0), np.float32(0.7089), np.float32(0.5561), np.float32(0.7198), np.float32(0.5509), np.float32(0.6731), np.float32(0.5877), np.float32(0.8057), np.float32(0.7899), np.float32(0.4967), np.float32(0.4759), np.float32(0.0), np.float32(0.0), np.float32(0.5948), np.float32(0.4161), np.float32(0.0), np.float32(0.0), np.float32(0.7538), np.float32(0.6861), np.float32(0.8533), np.float32(0.8619)] 
+2026-10-17 03:46:30.919897: Epoch time: 227.99 s 
+2026-10-17 03:46:31.816780:  
+2026-10-17 03:46:31.817390: Epoch 140 
+2026-10-17 03:46:31.817824: Current learning rate: 0.00873 
+2026-10-17 03:50:12.414771: train_loss -0.6255 
+2026-10-17 03:50:12.415376: val_loss -0.6207 
+2026-10-17 03:50:12.415905: Pseudo dice [np.float32(0.8856), np.float32(0.7609), np.float32(0.7207), np.float32(0.8692), np.float32(0.7011), np.float32(0.8463), np.float32(0.7841), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7448), np.float32(0.7226), np.float32(0.5987), np.float32(0.6271), np.float32(0.0), np.float32(0.0), np.float32(0.7099), np.float32(0.6184), np.float32(0.7262), np.float32(0.579), np.float32(0.6553), np.float32(0.6113), np.float32(0.7468), np.float32(0.7775), np.float32(0.5419), np.float32(0.5094), np.float32(0.0), np.float32(0.0), np.float32(0.2991), np.float32(0.5131), np.float32(0.0), np.float32(0.0), np.float32(0.7411), np.float32(0.6715), np.float32(0.8478), np.float32(0.834)] 
+2026-10-17 03:50:12.416870: Epoch time: 220.6 s 
+2026-10-17 03:50:13.237322:  
+2026-10-17 03:50:13.238087: Epoch 141 
+2026-10-17 03:50:13.238517: Current learning rate: 0.00872 
+2026-10-17 03:54:20.505139: train_loss -0.6545 
+2026-10-17 03:54:20.505918: val_loss -0.611 
+2026-10-17 03:54:20.506528: Pseudo dice [np.float32(0.8508), np.float32(0.7816), np.float32(0.7143), np.float32(0.8324), np.float32(0.7151), np.float32(0.8327), np.float32(0.7761), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.7493), np.float32(0.7362), np.float32(0.6852), np.float32(0.7062), np.float32(0.0), np.float32(0.0), np.float32(0.7567), np.float32(0.6176), np.float32(0.7339), np.float32(0.6033), np.float32(0.6579), np.float32(0.5926), np.float32(0.682), np.float32(0.7359), np.float32(0.467), np.float32(0.4233), np.float32(0.0), np.float32(0.0), np.float32(0.3483), np.float32(0.5087), np.float32(0.0), np.float32(0.0), np.float32(0.729), np.float32(0.7286), np.float32(0.8133), np.float32(0.7773)] 
+2026-10-17 03:54:20.507240: Epoch time: 247.27 s 
+2026-10-17 03:54:20.507609: Yayy! New best EMA pseudo Dice: 0.397599995136261 
+2026-10-17 03:54:23.055382:  
+2026-10-17 03:54:23.055777: Epoch 142 
+2026-10-17 03:54:23.056092: Current learning rate: 0.00871 
+2026-10-17 03:58:16.191099: train_loss -0.6496 
+2026-10-17 03:58:16.191939: val_loss -0.4378 
+2026-10-17 03:58:16.192505: Pseudo dice [np.float32(0.8644), np.float32(0.4859), np.float32(0.4748), np.float32(0.5762), np.float32(0.4734), np.float32(0.4425), np.float32(0.4324), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4562), np.float32(0.428), np.float32(0.5128), np.float32(0.4916), np.float32(0.0), np.float32(0.0), np.float32(0.5195), np.float32(0.3555), np.float32(0.4302), np.float32(0.3687), np.float32(0.2946), np.float32(0.2792), np.float32(0.3862), np.float32(0.4227), np.float32(0.3362), np.float32(0.244), np.float32(0.0), np.float32(0.0), np.float32(0.1845), np.float32(0.3897), np.float32(0.0), np.float32(0.0), np.float32(0.4887), np.float32(0.3062), np.float32(0.4939), np.float32(0.4509)] 
+2026-10-17 03:58:16.193166: Epoch time: 233.14 s 
+2026-10-17 03:58:17.043735:  
+2026-10-17 03:58:17.044372: Epoch 143 
+2026-10-17 03:58:17.044648: Current learning rate: 0.0087 
+2026-10-17 04:02:11.455878: train_loss -0.6515 
+2026-10-17 04:02:11.456463: val_loss -0.3908 
+2026-10-17 04:02:11.457216: Pseudo dice [np.float32(0.8642), np.float32(0.4174), np.float32(0.3858), np.float32(0.4326), np.float32(0.4009), np.float32(0.3812), np.float32(0.3925), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3851), np.float32(0.3253), np.float32(0.3672), np.float32(0.1893), np.float32(0.0), np.float32(0.0), np.float32(0.4321), np.float32(0.3725), np.float32(0.3561), np.float32(0.2828), np.float32(0.154), np.float32(0.2359), np.float32(0.4927), np.float32(0.5654), np.float32(0.2013), np.float32(0.236), np.float32(0.0), np.float32(0.0), np.float32(0.0419), np.float32(0.3485), np.float32(0.0), np.float32(0.0), np.float32(0.2976), np.float32(0.2507), np.float32(0.3927), np.float32(0.3932)] 
+2026-10-17 04:02:11.457854: Epoch time: 234.41 s 
+2026-10-17 04:02:12.334154:  
+2026-10-17 04:02:12.334647: Epoch 144 
+2026-10-17 04:02:12.334980: Current learning rate: 0.00869 
+2026-10-17 04:05:58.278008: train_loss -0.6709 
+2026-10-17 04:05:58.278969: val_loss -0.5441 
+2026-10-17 04:05:58.279304: Pseudo dice [np.float32(0.863), np.float32(0.6343), np.float32(0.6022), np.float32(0.6608), np.float32(0.561), np.float32(0.6323), np.float32(0.6401), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5235), np.float32(0.5482), np.float32(0.4778), np.float32(0.4813), np.float32(0.0), np.float32(0.0), np.float32(0.5728), np.float32(0.4524), np.float32(0.579), np.float32(0.4978), np.float32(0.4953), np.float32(0.4169), np.float32(0.7731), np.float32(0.7596), np.float32(0.3313), np.float32(0.3976), np.float32(0.0), np.float32(0.0), np.float32(0.2244), np.float32(0.4309), np.float32(0.0), np.float32(0.0), np.float32(0.4981), np.float32(0.4715), np.float32(0.6577), np.float32(0.6603)] 
+2026-10-17 04:05:58.279834: Epoch time: 225.95 s 
+2026-10-17 04:05:59.119201:  
+2026-10-17 04:05:59.119823: Epoch 145 
+2026-10-17 04:05:59.120164: Current learning rate: 0.00868 
+2026-10-17 04:09:45.781594: train_loss -0.6381 
+2026-10-17 04:09:45.782646: val_loss -0.2222 
+2026-10-17 04:09:45.783367: Pseudo dice [np.float32(0.8741), np.float32(0.094), np.float32(0.1366), np.float32(0.2869), np.float32(0.127), np.float32(0.0964), np.float32(0.118), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3002), np.float32(0.1134), np.float32(0.2006), np.float32(0.1147), np.float32(0.0), np.float32(0.0), np.float32(0.1344), np.float32(0.1278), np.float32(0.1048), np.float32(0.1008), np.float32(0.0969), np.float32(0.0485), np.float32(0.223), np.float32(0.1191), np.float32(0.0486), np.float32(0.0118), np.float32(0.0), np.float32(0.0), np.float32(0.0401), np.float32(0.0386), np.float32(0.0), np.float32(0.0), np.float32(0.0964), np.float32(0.0638), np.float32(0.1522), np.float32(0.0963)] 
+2026-10-17 04:09:45.783669: Epoch time: 226.66 s 
+2026-10-17 04:09:46.950009:  
+2026-10-17 04:09:46.950722: Epoch 146 
+2026-10-17 04:09:46.951101: Current learning rate: 0.00868 
+2026-10-17 04:13:38.120105: train_loss -0.576 
+2026-10-17 04:13:38.120994: val_loss -0.3533 
+2026-10-17 04:13:38.121674: Pseudo dice [np.float32(0.8798), np.float32(0.1667), np.float32(0.3984), np.float32(0.2511), np.float32(0.1796), np.float32(0.4011), np.float32(0.4456), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2935), np.float32(0.3154), np.float32(0.4234), np.float32(0.1873), np.float32(0.0), np.float32(0.0), np.float32(0.2701), np.float32(0.2458), np.float32(0.4599), np.float32(0.3042), np.float32(0.3414), np.float32(0.2949), np.float32(0.4709), np.float32(0.2916), np.float32(0.106), np.float32(0.1916), np.float32(0.0), np.float32(0.0), np.float32(0.2522), np.float32(0.2231), np.float32(0.0), np.float32(0.0), np.float32(0.2828), np.float32(0.3555), np.float32(0.34), np.float32(0.3275)] 
+2026-10-17 04:13:38.122231: Epoch time: 231.17 s 
+2026-10-17 04:13:39.021179:  
+2026-10-17 04:13:39.021903: Epoch 147 
+2026-10-17 04:13:39.022252: Current learning rate: 0.00867 
+2026-10-17 04:17:42.624246: train_loss -0.6155 
+2026-10-17 04:17:42.624887: val_loss -0.5327 
+2026-10-17 04:17:42.625378: Pseudo dice [np.float32(0.8839), np.float32(0.6899), np.float32(0.5985), np.float32(0.7106), np.float32(0.5753), np.float32(0.6854), np.float32(0.6113), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6498), np.float32(0.6347), np.float32(0.6463), np.float32(0.6378), np.float32(0.0), np.float32(0.0), np.float32(0.5861), np.float32(0.4865), np.float32(0.6022), np.float32(0.5151), np.float32(0.5676), np.float32(0.4471), np.float32(0.7774), np.float32(0.7359), np.float32(0.4795), np.float32(0.485), np.float32(0.0), np.float32(0.0), np.float32(0.3383), np.float32(0.2969), np.float32(0.0), np.float32(0.0), np.float32(0.6125), np.float32(0.5664), np.float32(0.6998), np.float32(0.705)] 
+2026-10-17 04:17:42.626009: Epoch time: 243.61 s 
+2026-10-17 04:17:43.505924:  
+2026-10-17 04:17:43.506557: Epoch 148 
+2026-10-17 04:17:43.506899: Current learning rate: 0.00866 
+2026-10-17 04:21:26.804419: train_loss -0.605 
+2026-10-17 04:21:26.805037: val_loss -0.4169 
+2026-10-17 04:21:26.805555: Pseudo dice [np.float32(0.8764), np.float32(0.5334), np.float32(0.4973), np.float32(0.3288), np.float32(0.4084), np.float32(0.4169), np.float32(0.4326), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3599), np.float32(0.3641), np.float32(0.4682), np.float32(0.458), np.float32(0.0), np.float32(0.0), np.float32(0.4089), np.float32(0.3349), np.float32(0.397), np.float32(0.425), np.float32(0.5113), np.float32(0.4557), np.float32(0.6441), np.float32(0.6053), np.float32(0.3593), np.float32(0.2156), np.float32(0.0), np.float32(0.0), np.float32(0.4246), np.float32(0.2137), np.float32(0.0), np.float32(0.0), np.float32(0.427), np.float32(0.4729), np.float32(0.543), np.float32(0.5308)] 
+2026-10-17 04:21:26.805921: Epoch time: 223.3 s 
+2026-10-17 04:21:27.688780:  
+2026-10-17 04:21:27.689403: Epoch 149 
+2026-10-17 04:21:27.690133: Current learning rate: 0.00865 
+2026-10-17 04:25:26.012114: train_loss -0.6254 
+2026-10-17 04:25:26.012712: val_loss -0.5368 
+2026-10-17 04:25:26.013291: Pseudo dice [np.float32(0.8482), np.float32(0.6618), np.float32(0.5701), np.float32(0.6782), np.float32(0.5897), np.float32(0.6613), np.float32(0.6094), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6019), np.float32(0.6296), np.float32(0.5063), np.float32(0.5222), np.float32(0.0), np.float32(0.0), np.float32(0.5972), np.float32(0.4707), np.float32(0.5564), np.float32(0.4714), np.float32(0.4954), np.float32(0.4027), np.float32(0.7963), np.float32(0.7959), np.float32(0.4564), np.float32(0.4601), np.float32(0.0), np.float32(0.0), np.float32(0.1661), np.float32(0.4342), np.float32(0.0), np.float32(0.0), np.float32(0.4694), np.float32(0.405), np.float32(0.6904), np.float32(0.6984)] 
+2026-10-17 04:25:26.014163: Epoch time: 238.33 s 
+2026-10-17 04:25:27.817893:  
+2026-10-17 04:25:27.818701: Epoch 150 
+2026-10-17 04:25:27.836458: Current learning rate: 0.00864 
+2026-10-17 04:29:12.356170: train_loss -0.6513 
+2026-10-17 04:29:12.357105: val_loss -0.3932 
+2026-10-17 04:29:12.357561: Pseudo dice [np.float32(0.8663), np.float32(0.3722), np.float32(0.3537), np.float32(0.4365), np.float32(0.3208), np.float32(0.4321), np.float32(0.3786), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4575), np.float32(0.45), np.float32(0.4422), np.float32(0.419), np.float32(0.0), np.float32(0.0), np.float32(0.3618), np.float32(0.2697), np.float32(0.3888), np.float32(0.356), np.float32(0.3269), np.float32(0.2796), np.float32(0.3443), np.float32(0.2831), np.float32(0.3227), np.float32(0.2711), np.float32(0.0), np.float32(0.0), np.float32(0.3972), np.float32(0.2441), np.float32(0.0), np.float32(0.0), np.float32(0.4129), np.float32(0.4008), np.float32(0.4145), np.float32(0.4099)] 
+2026-10-17 04:29:12.357994: Epoch time: 224.54 s 
+2026-10-17 04:29:13.255742:  
+2026-10-17 04:29:13.256471: Epoch 151 
+2026-10-17 04:29:13.256914: Current learning rate: 0.00863 
+2026-10-17 04:32:50.000053: train_loss -0.6138 
+2026-10-17 04:32:50.000580: val_loss -0.5102 
+2026-10-17 04:32:50.001209: Pseudo dice [np.float32(0.8236), np.float32(0.5762), np.float32(0.5396), np.float32(0.6343), np.float32(0.5663), np.float32(0.5969), np.float32(0.6591), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5353), np.float32(0.531), np.float32(0.5822), np.float32(0.6069), np.float32(0.0), np.float32(0.0), np.float32(0.6027), np.float32(0.4526), np.float32(0.587), np.float32(0.4897), np.float32(0.4325), np.float32(0.3675), np.float32(0.3594), np.float32(0.5754), np.float32(0.5666), np.float32(0.4977), np.float32(0.0), np.float32(0.0), np.float32(0.4223), np.float32(0.2432), np.float32(0.0), np.float32(0.0), np.float32(0.5769), np.float32(0.5711), np.float32(0.6668), np.float32(0.6721)] 
+2026-10-17 04:32:50.002209: Epoch time: 216.75 s 
+2026-10-17 04:32:50.840782:  
+2026-10-17 04:32:50.841624: Epoch 152 
+2026-10-17 04:32:50.842010: Current learning rate: 0.00862 
+2026-10-17 04:36:44.502914: train_loss -0.5927 
+2026-10-17 04:36:44.503618: val_loss -0.3713 
+2026-10-17 04:36:44.504367: Pseudo dice [np.float32(0.861), np.float32(0.384), np.float32(0.3711), np.float32(0.4692), np.float32(0.4341), np.float32(0.4186), np.float32(0.4308), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3845), np.float32(0.3285), np.float32(0.311), np.float32(0.4148), np.float32(0.0), np.float32(0.0), np.float32(0.3397), np.float32(0.2158), np.float32(0.393), np.float32(0.3321), np.float32(0.1396), np.float32(0.1995), np.float32(0.4542), np.float32(0.3141), np.float32(0.1578), np.float32(0.1132), np.float32(0.0), np.float32(0.0), np.float32(0.328), np.float32(0.1438), np.float32(0.0), np.float32(0.0), np.float32(0.4552), np.float32(0.4629), np.float32(0.3982), np.float32(0.3803)] 
+2026-10-17 04:36:44.505144: Epoch time: 233.66 s 
+2026-10-17 04:36:45.358149:  
+2026-10-17 04:36:45.358844: Epoch 153 
+2026-10-17 04:36:45.359254: Current learning rate: 0.00861 
+2026-10-17 04:40:21.002395: train_loss -0.6157 
+2026-10-17 04:40:21.003071: val_loss -0.4185 
+2026-10-17 04:40:21.003749: Pseudo dice [np.float32(0.8591), np.float32(0.3827), np.float32(0.4064), np.float32(0.5765), np.float32(0.4588), np.float32(0.5511), np.float32(0.4745), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5294), np.float32(0.511), np.float32(0.4336), np.float32(0.4202), np.float32(0.0), np.float32(0.0), np.float32(0.482), np.float32(0.3412), np.float32(0.472), np.float32(0.3237), np.float32(0.3161), np.float32(0.2458), np.float32(0.5647), np.float32(0.5781), np.float32(0.4181), np.float32(0.2987), np.float32(0.0), np.float32(0.0), np.float32(0.1521), np.float32(0.3759), np.float32(0.0), np.float32(0.0), np.float32(0.4705), np.float32(0.427), np.float32(0.4902), np.float32(0.4546)] 
+2026-10-17 04:40:21.004924: Epoch time: 215.65 s 
+2026-10-17 04:40:21.933625:  
+2026-10-17 04:40:21.934255: Epoch 154 
+2026-10-17 04:40:21.934655: Current learning rate: 0.0086 
+2026-10-17 04:44:14.086832: train_loss -0.5929 
+2026-10-17 04:44:14.087758: val_loss -0.5931 
+2026-10-17 04:44:14.088498: Pseudo dice [np.float32(0.8869), np.float32(0.7298), np.float32(0.6713), np.float32(0.8074), np.float32(0.7057), np.float32(0.8149), np.float32(0.7352), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.655), np.float32(0.6556), np.float32(0.5555), np.float32(0.641), np.float32(0.0), np.float32(0.0), np.float32(0.721), np.float32(0.5301), np.float32(0.6751), np.float32(0.5483), np.float32(0.5932), np.float32(0.4811), np.float32(0.8347), np.float32(0.797), np.float32(0.5762), np.float32(0.5616), np.float32(0.0), np.float32(0.0), np.float32(0.5167), np.float32(0.2806), np.float32(0.0), np.float32(0.0), np.float32(0.6737), np.float32(0.6543), np.float32(0.8033), np.float32(0.8298)] 
+2026-10-17 04:44:14.088782: Epoch time: 232.16 s 
+2026-10-17 04:44:14.962329:  
+2026-10-17 04:44:14.963303: Epoch 155 
+2026-10-17 04:44:14.963734: Current learning rate: 0.00859 
+2026-10-17 04:47:58.790549: train_loss -0.5997 
+2026-10-17 04:47:58.791238: val_loss -0.468 
+2026-10-17 04:47:58.791780: Pseudo dice [np.float32(0.8389), np.float32(0.4316), np.float32(0.4717), np.float32(0.5843), np.float32(0.4464), np.float32(0.5646), np.float32(0.4766), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5048), np.float32(0.4517), np.float32(0.4565), np.float32(0.3887), np.float32(0.0), np.float32(0.0), np.float32(0.5112), np.float32(0.3688), np.float32(0.4968), np.float32(0.3653), np.float32(0.2286), np.float32(0.3371), np.float32(0.3546), np.float32(0.3495), np.float32(0.2933), np.float32(0.3025), np.float32(0.0), np.float32(0.0), np.float32(0.3224), np.float32(0.2295), np.float32(0.0), np.float32(0.0), np.float32(0.5586), np.float32(0.3672), np.float32(0.5517), np.float32(0.5048)] 
+2026-10-17 04:47:58.792239: Epoch time: 223.83 s 
+2026-10-17 04:47:59.691200:  
+2026-10-17 04:47:59.692131: Epoch 156 
+2026-10-17 04:47:59.692384: Current learning rate: 0.00858 
+2026-10-17 04:51:31.534376: train_loss -0.6395 
+2026-10-17 04:51:31.534962: val_loss -0.5219 
+2026-10-17 04:51:31.535840: Pseudo dice [np.float32(0.8031), np.float32(0.6086), np.float32(0.5589), np.float32(0.6877), np.float32(0.5559), np.float32(0.7107), np.float32(0.583), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.6119), np.float32(0.5829), np.float32(0.5162), np.float32(0.4472), np.float32(0.0), np.float32(0.0), np.float32(0.5864), np.float32(0.4862), np.float32(0.6087), np.float32(0.4513), np.float32(0.6072), np.float32(0.5694), np.float32(0.5754), np.float32(0.6986), np.float32(0.4839), np.float32(0.5187), np.float32(0.0), np.float32(0.0), np.float32(0.4451), np.float32(0.4014), np.float32(0.0), np.float32(0.0), np.float32(0.5602), np.float32(0.5991), np.float32(0.7017), np.float32(0.7034)] 
+2026-10-17 04:51:31.536283: Epoch time: 211.85 s 
+2026-10-17 04:51:32.386047:  
+2026-10-17 04:51:32.386744: Epoch 157 
+2026-10-17 04:51:32.387176: Current learning rate: 0.00858 
+2026-10-17 04:55:07.256208: train_loss -0.6393 
+2026-10-17 04:55:07.256811: val_loss -0.2174 
+2026-10-17 04:55:07.257370: Pseudo dice [np.float32(0.8639), np.float32(0.1142), np.float32(0.0817), np.float32(0.1076), np.float32(0.0744), np.float32(0.0836), np.float32(0.0648), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.1531), np.float32(0.1172), np.float32(0.15), np.float32(0.1611), np.float32(0.0), np.float32(0.0), np.float32(0.1227), np.float32(0.1282), np.float32(0.0507), np.float32(0.0845), np.float32(0.0711), np.float32(0.0789), np.float32(0.2368), np.float32(0.1623), np.float32(0.2114), np.float32(0.1887), np.float32(0.0), np.float32(0.0), np.float32(0.0377), np.float32(0.0213), np.float32(0.0), np.float32(0.0), np.float32(0.0817), np.float32(0.0387), np.float32(0.0832), np.float32(0.1183)] 
+2026-10-17 04:55:07.257941: Epoch time: 214.87 s 
+2026-10-17 04:55:08.150046:  
+2026-10-17 04:55:08.150854: Epoch 158 
+2026-10-17 04:55:08.151307: Current learning rate: 0.00857 
+2026-10-17 04:59:10.216067: train_loss -0.5491 
+2026-10-17 04:59:10.216598: val_loss -0.309 
+2026-10-17 04:59:10.217065: Pseudo dice [np.float32(0.7973), np.float32(0.1985), np.float32(0.2433), np.float32(0.2514), np.float32(0.0986), np.float32(0.225), np.float32(0.1961), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2701), np.float32(0.2831), np.float32(0.3693), np.float32(0.3486), np.float32(0.0), np.float32(0.0), np.float32(0.1575), np.float32(0.145), np.float32(0.2517), np.float32(0.2545), np.float32(0.1243), np.float32(0.2267), np.float32(0.3543), np.float32(0.148), np.float32(0.2514), np.float32(0.2404), np.float32(0.0), np.float32(0.0), np.float32(0.0635), np.float32(0.1812), np.float32(0.0), np.float32(0.0), np.float32(0.1976), np.float32(0.1631), np.float32(0.276), np.float32(0.231)] 
+2026-10-17 04:59:10.218064: Epoch time: 242.07 s 
+2026-10-17 04:59:11.068480:  
+2026-10-17 04:59:11.069096: Epoch 159 
+2026-10-17 04:59:11.069623: Current learning rate: 0.00856 
+2026-10-17 05:02:46.263572: train_loss -0.5886 
+2026-10-17 05:02:46.264401: val_loss -0.5509 
+2026-10-17 05:02:46.265162: Pseudo dice [np.float32(0.8743), np.float32(0.6895), np.float32(0.6425), np.float32(0.6838), np.float32(0.584), np.float32(0.6777), np.float32(0.6177), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.569), np.float32(0.562), np.float32(0.5324), np.float32(0.6005), np.float32(0.0), np.float32(0.0), np.float32(0.5577), np.float32(0.4251), np.float32(0.5553), np.float32(0.4381), np.float32(0.4866), np.float32(0.3752), np.float32(0.7107), np.float32(0.739), np.float32(0.5094), np.float32(0.5308), np.float32(0.0), np.float32(0.0), np.float32(0.2039), np.float32(0.3781), np.float32(0.0), np.float32(0.0), np.float32(0.5299), np.float32(0.5034), np.float32(0.6426), np.float32(0.6482)] 
+2026-10-17 05:02:46.265541: Epoch time: 215.2 s 
+2026-10-17 05:02:47.077963:  
+2026-10-17 05:02:47.079036: Epoch 160 
+2026-10-17 05:02:47.079371: Current learning rate: 0.00855 
+2026-10-17 05:06:41.364719: train_loss -0.6247 
+2026-10-17 05:06:41.365519: val_loss -0.4738 
+2026-10-17 05:06:41.366132: Pseudo dice [np.float32(0.8494), np.float32(0.5105), np.float32(0.4755), np.float32(0.5716), np.float32(0.4963), np.float32(0.5094), np.float32(0.5138), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4436), np.float32(0.4857), np.float32(0.4788), np.float32(0.4911), np.float32(0.0), np.float32(0.0), np.float32(0.5332), np.float32(0.402), np.float32(0.4838), np.float32(0.3927), np.float32(0.369), np.float32(0.3655), np.float32(0.4429), np.float32(0.543), np.float32(0.3839), np.float32(0.4161), np.float32(0.0), np.float32(0.0), np.float32(0.3539), np.float32(0.4518), np.float32(0.0), np.float32(0.0), np.float32(0.4017), np.float32(0.3629), np.float32(0.5148), np.float32(0.4549)] 
+2026-10-17 05:06:41.366519: Epoch time: 234.29 s 
+2026-10-17 05:06:42.366387:  
+2026-10-17 05:06:42.367055: Epoch 161 
+2026-10-17 05:06:42.367490: Current learning rate: 0.00854 
+2026-10-17 05:10:20.082602: train_loss -0.6499 
+2026-10-17 05:10:20.083441: val_loss -0.4289 
+2026-10-17 05:10:20.084051: Pseudo dice [np.float32(0.8801), np.float32(0.3544), np.float32(0.3854), np.float32(0.5029), np.float32(0.3938), np.float32(0.4983), np.float32(0.4207), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3809), np.float32(0.3873), np.float32(0.4536), np.float32(0.405), np.float32(0.0), np.float32(0.0), np.float32(0.3733), np.float32(0.2552), np.float32(0.3585), np.float32(0.3136), np.float32(0.2889), np.float32(0.3041), np.float32(0.2355), np.float32(0.3804), np.float32(0.3322), np.float32(0.2305), np.float32(0.0), np.float32(0.0), np.float32(0.3502), np.float32(0.3064), np.float32(0.0), np.float32(0.0), np.float32(0.4218), np.float32(0.4238), np.float32(0.4287), np.float32(0.412)] 
+2026-10-17 05:10:20.084551: Epoch time: 217.72 s 
+2026-10-17 05:10:21.316526:  
+2026-10-17 05:10:21.317322: Epoch 162 
+2026-10-17 05:10:21.317752: Current learning rate: 0.00853 
+2026-10-17 05:14:14.648985: train_loss -0.6725 
+2026-10-17 05:14:14.649846: val_loss -0.4835 
+2026-10-17 05:14:14.650372: Pseudo dice [np.float32(0.8255), np.float32(0.5302), np.float32(0.5235), np.float32(0.629), np.float32(0.5281), np.float32(0.6579), np.float32(0.5569), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5009), np.float32(0.471), np.float32(0.5066), np.float32(0.5525), np.float32(0.0), np.float32(0.0), np.float32(0.5771), np.float32(0.4368), np.float32(0.523), np.float32(0.4369), np.float32(0.5931), np.float32(0.4684), np.float32(0.5083), np.float32(0.4857), np.float32(0.3461), np.float32(0.3105), np.float32(0.0), np.float32(0.0), np.float32(0.5399), np.float32(0.306), np.float32(0.0), np.float32(0.0), np.float32(0.5874), np.float32(0.5778), np.float32(0.6022), np.float32(0.5867)] 
+2026-10-17 05:14:14.650742: Epoch time: 233.33 s 
+2026-10-17 05:14:15.503582:  
+2026-10-17 05:14:15.504145: Epoch 163 
+2026-10-17 05:14:15.504616: Current learning rate: 0.00852 
+2026-10-17 05:17:58.477411: train_loss -0.6718 
+2026-10-17 05:17:58.478472: val_loss -0.376 
+2026-10-17 05:17:58.479069: Pseudo dice [np.float32(0.8897), np.float32(0.3265), np.float32(0.3768), np.float32(0.489), np.float32(0.3583), np.float32(0.4508), np.float32(0.3699), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3948), np.float32(0.3649), np.float32(0.4785), np.float32(0.4723), np.float32(0.0), np.float32(0.0), np.float32(0.3162), np.float32(0.2323), np.float32(0.352), np.float32(0.2912), np.float32(0.4063), np.float32(0.2985), np.float32(0.1748), np.float32(0.2008), np.float32(0.2633), np.float32(0.1593), np.float32(0.0), np.float32(0.0), np.float32(0.3506), np.float32(0.216), np.float32(0.0), np.float32(0.0), np.float32(0.4078), np.float32(0.3706), np.float32(0.341), np.float32(0.3155)] 
+2026-10-17 05:17:58.479542: Epoch time: 222.98 s 
+2026-10-17 05:17:59.400130:  
+2026-10-17 05:17:59.401084: Epoch 164 
+2026-10-17 05:17:59.401477: Current learning rate: 0.00851 
+2026-10-17 05:22:06.735081: train_loss -0.6747 
+2026-10-17 05:22:06.736020: val_loss -0.3046 
+2026-10-17 05:22:06.736680: Pseudo dice [np.float32(0.8786), np.float32(0.2136), np.float32(0.2502), np.float32(0.329), np.float32(0.2181), np.float32(0.2598), np.float32(0.2305), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.2401), np.float32(0.2116), np.float32(0.2732), np.float32(0.2569), np.float32(0.0), np.float32(0.0), np.float32(0.2435), np.float32(0.1496), np.float32(0.2008), np.float32(0.1658), np.float32(0.2244), np.float32(0.2348), np.float32(0.1993), np.float32(0.3649), np.float32(0.1961), np.float32(0.0782), np.float32(0.0), np.float32(0.0), np.float32(0.2982), np.float32(0.2593), np.float32(0.0), np.float32(0.0), np.float32(0.2129), np.float32(0.2345), np.float32(0.304), np.float32(0.2157)] 
+2026-10-17 05:22:06.737077: Epoch time: 247.34 s 
+2026-10-17 05:22:07.620002:  
+2026-10-17 05:22:07.620675: Epoch 165 
+2026-10-17 05:22:07.620986: Current learning rate: 0.0085 
+2026-10-17 05:26:18.142131: train_loss -0.6796 
+2026-10-17 05:26:18.142848: val_loss -0.5006 
+2026-10-17 05:26:18.143598: Pseudo dice [np.float32(0.8647), np.float32(0.5349), np.float32(0.5408), np.float32(0.6482), np.float32(0.5638), np.float32(0.6492), np.float32(0.5468), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.5158), np.float32(0.5181), np.float32(0.4859), np.float32(0.5309), np.float32(0.0), np.float32(0.0), np.float32(0.5693), np.float32(0.4695), np.float32(0.5058), np.float32(0.4211), np.float32(0.5739), np.float32(0.4234), np.float32(0.5727), np.float32(0.599), np.float32(0.3663), np.float32(0.2817), np.float32(0.0), np.float32(0.0), np.float32(0.4438), np.float32(0.3103), np.float32(0.0), np.float32(0.0), np.float32(0.5776), np.float32(0.5892), np.float32(0.6167), np.float32(0.6089)] 
+2026-10-17 05:26:18.144283: Epoch time: 250.52 s 
+2026-10-17 05:26:19.022645:  
+2026-10-17 05:26:19.023467: Epoch 166 
+2026-10-17 05:26:19.023915: Current learning rate: 0.00849 
+2026-10-17 05:29:47.798921: train_loss -0.6704 
+2026-10-17 05:29:47.799719: val_loss -0.3983 
+2026-10-17 05:29:47.800213: Pseudo dice [np.float32(0.8857), np.float32(0.4037), np.float32(0.4059), np.float32(0.4873), np.float32(0.3181), np.float32(0.5227), np.float32(0.3535), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.4659), np.float32(0.4149), np.float32(0.4917), np.float32(0.398), np.float32(0.0), np.float32(0.0), np.float32(0.3743), np.float32(0.3337), np.float32(0.3065), np.float32(0.2161), np.float32(0.3013), np.float32(0.3433), np.float32(0.2565), np.float32(0.4494), np.float32(0.3139), np.float32(0.2777), np.float32(0.0), np.float32(0.0), np.float32(0.3947), np.float32(0.405), np.float32(0.0), np.float32(0.0), np.float32(0.3673), np.float32(0.3492), np.float32(0.3752), np.float32(0.379)] 
+2026-10-17 05:29:47.800620: Epoch time: 208.78 s 
+2026-10-17 05:29:48.594256:  
+2026-10-17 05:29:48.594957: Epoch 167 
+2026-10-17 05:29:48.595362: Current learning rate: 0.00848 
diff --git a/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/plans.json b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/plans.json
new file mode 100644
index 0000000000000000000000000000000000000000..5e91dc1c79ce056bedda5be7fd435ccfb89f8dd4
--- /dev/null
+++ b/nnUNet_Large/fold_1/nnUNet_results/Dataset001_CT/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/plans.json
@@ -0,0 +1,356 @@
+{
+    "dataset_name": "Dataset001_CT",
+    "plans_name": "nnUNetResEncUNetLPlans",
+    "original_median_spacing_after_transp": [
+        0.70001220703125,
+        0.482421875,
+        0.482421875
+    ],
+    "original_median_shape_after_transp": [
+        206,
+        365,
+        286
+    ],
+    "image_reader_writer": "SimpleITKIO",
+    "transpose_forward": [
+        0,
+        1,
+        2
+    ],
+    "transpose_backward": [
+        0,
+        1,
+        2
+    ],
+    "configurations": {
+        "2d": {
+            "data_identifier": "nnUNetPlans_2d",
+            "preprocessor_name": "DefaultPreprocessor",
+            "batch_size": 70,
+            "patch_size": [
+                384,
+                320
+            ],
+            "median_image_size_in_voxels": [
+                352.0,
+                281.0
+            ],
+            "spacing": [
+                0.482421875,
+                0.482421875
+            ],
+            "normalization_schemes": [
+                "CTNormalization"
+            ],
+            "use_mask_for_norm": [
+                false
+            ],
+            "resampling_fn_data": "resample_data_or_seg_to_shape",
+            "resampling_fn_seg": "resample_data_or_seg_to_shape",
+            "resampling_fn_data_kwargs": {
+                "is_seg": false,
+                "order": 3,
+                "order_z": 0,
+                "force_separate_z": null
+            },
+            "resampling_fn_seg_kwargs": {
+                "is_seg": true,
+                "order": 1,
+                "order_z": 0,
+                "force_separate_z": null
+            },
+            "resampling_fn_probabilities": "resample_data_or_seg_to_shape",
+            "resampling_fn_probabilities_kwargs": {
+                "is_seg": false,
+                "order": 1,
+                "order_z": 0,
+                "force_separate_z": null
+            },
+            "architecture": {
+                "network_class_name": "dynamic_network_architectures.architectures.unet.ResidualEncoderUNet",
+                "arch_kwargs": {
+                    "n_stages": 7,
+                    "features_per_stage": [
+                        32,
+                        64,
+                        128,
+                        256,
+                        512,
+                        512,
+                        512
+                    ],
+                    "conv_op": "torch.nn.modules.conv.Conv2d",
+                    "kernel_sizes": [
+                        [
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3
+                        ]
+                    ],
+                    "strides": [
+                        [
+                            1,
+                            1
+                        ],
+                        [
+                            2,
+                            2
+                        ],
+                        [
+                            2,
+                            2
+                        ],
+                        [
+                            2,
+                            2
+                        ],
+                        [
+                            2,
+                            2
+                        ],
+                        [
+                            2,
+                            2
+                        ],
+                        [
+                            2,
+                            2
+                        ]
+                    ],
+                    "n_blocks_per_stage": [
+                        1,
+                        3,
+                        4,
+                        6,
+                        6,
+                        6,
+                        6
+                    ],
+                    "n_conv_per_stage_decoder": [
+                        1,
+                        1,
+                        1,
+                        1,
+                        1,
+                        1
+                    ],
+                    "conv_bias": true,
+                    "norm_op": "torch.nn.modules.instancenorm.InstanceNorm2d",
+                    "norm_op_kwargs": {
+                        "eps": 1e-05,
+                        "affine": true
+                    },
+                    "dropout_op": null,
+                    "dropout_op_kwargs": null,
+                    "nonlin": "torch.nn.LeakyReLU",
+                    "nonlin_kwargs": {
+                        "inplace": true
+                    }
+                },
+                "_kw_requires_import": [
+                    "conv_op",
+                    "norm_op",
+                    "dropout_op",
+                    "nonlin"
+                ]
+            },
+            "batch_dice": true
+        },
+        "3d_fullres": {
+            "data_identifier": "nnUNetPlans_3d_fullres",
+            "preprocessor_name": "DefaultPreprocessor",
+            "batch_size": 2,
+            "patch_size": [
+                128,
+                256,
+                192
+            ],
+            "median_image_size_in_voxels": [
+                214.0,
+                352.0,
+                281.0
+            ],
+            "spacing": [
+                0.70001220703125,
+                0.482421875,
+                0.482421875
+            ],
+            "normalization_schemes": [
+                "CTNormalization"
+            ],
+            "use_mask_for_norm": [
+                false
+            ],
+            "resampling_fn_data": "resample_data_or_seg_to_shape",
+            "resampling_fn_seg": "resample_data_or_seg_to_shape",
+            "resampling_fn_data_kwargs": {
+                "is_seg": false,
+                "order": 3,
+                "order_z": 0,
+                "force_separate_z": null
+            },
+            "resampling_fn_seg_kwargs": {
+                "is_seg": true,
+                "order": 1,
+                "order_z": 0,
+                "force_separate_z": null
+            },
+            "resampling_fn_probabilities": "resample_data_or_seg_to_shape",
+            "resampling_fn_probabilities_kwargs": {
+                "is_seg": false,
+                "order": 1,
+                "order_z": 0,
+                "force_separate_z": null
+            },
+            "architecture": {
+                "network_class_name": "dynamic_network_architectures.architectures.unet.ResidualEncoderUNet",
+                "arch_kwargs": {
+                    "n_stages": 6,
+                    "features_per_stage": [
+                        32,
+                        64,
+                        128,
+                        256,
+                        320,
+                        320
+                    ],
+                    "conv_op": "torch.nn.modules.conv.Conv3d",
+                    "kernel_sizes": [
+                        [
+                            3,
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3,
+                            3
+                        ],
+                        [
+                            3,
+                            3,
+                            3
+                        ]
+                    ],
+                    "strides": [
+                        [
+                            1,
+                            1,
+                            1
+                        ],
+                        [
+                            2,
+                            2,
+                            2
+                        ],
+                        [
+                            2,
+                            2,
+                            2
+                        ],
+                        [
+                            2,
+                            2,
+                            2
+                        ],
+                        [
+                            2,
+                            2,
+                            2
+                        ],
+                        [
+                            2,
+                            2,
+                            2
+                        ]
+                    ],
+                    "n_blocks_per_stage": [
+                        1,
+                        3,
+                        4,
+                        6,
+                        6,
+                        6
+                    ],
+                    "n_conv_per_stage_decoder": [
+                        1,
+                        1,
+                        1,
+                        1,
+                        1
+                    ],
+                    "conv_bias": true,
+                    "norm_op": "torch.nn.modules.instancenorm.InstanceNorm3d",
+                    "norm_op_kwargs": {
+                        "eps": 1e-05,
+                        "affine": true
+                    },
+                    "dropout_op": null,
+                    "dropout_op_kwargs": null,
+                    "nonlin": "torch.nn.LeakyReLU",
+                    "nonlin_kwargs": {
+                        "inplace": true
+                    }
+                },
+                "_kw_requires_import": [
+                    "conv_op",
+                    "norm_op",
+                    "dropout_op",
+                    "nonlin"
+                ]
+            },
+            "batch_dice": false
+        }
+    },
+    "experiment_planner_used": "nnUNetPlannerResEncL",
+    "label_manager": "LabelManager",
+    "foreground_intensity_properties_per_channel": {
+        "0": {
+            "max": 1813.9854736328125,
+            "mean": 211.27015686035156,
+            "median": 192.95703125,
+            "min": -91.01630401611328,
+            "percentile_00_5": 66.00397491455078,
+            "percentile_99_5": 618.0081787109375,
+            "std": 107.36483764648438
+        }
+    }
+}
\ No newline at end of file